(c)2006CrashDataGroupInc
The ARC Network and Collision Safety Institute host the annual premier crash conference, the
This conference is designed to be topical, timely and a lasting learning experience for all involved.
Every June in Las Vegas, NV you have the opportunity to witness live crash testing first hand, attend presentations by recognized leaders in their field, and leave the conference with crash data in hand including the video, photos, copies of presentations, and related documents.
There is no better opportunity for training. Anywhere.
The ARC Network and Collision Safety Institute along with Vetronix, CDR Tool User’s Group and other sponsors host the newest conference and training opportunity, the
(c)2006CrashDataGroupInc
This is a comprehensive conference held in Texas in January/February to publish CDR System related information in the form of peer-reviewed papers, data compendium and the like. It covers legal issues, current legislation, user applications and future updates and developments.
www.crashconferences.com
fall 2006
of interest
From the Editor
NAPARS News
Organization News
My Turn at the Wheel Calendar of Events
PDOF: Principle Direction of Force
Deceleration Rates of Modern Passenger Vehicles During Straight Line Braking and Yaw Events
identification of Unusual Tire Marks at the Scene of a Motor Vehicle Collision
Motorcycle Crash Investigation
Simulations 101: Anatomy of a Simulation
Pictometry in Crash Scene Mapping
Making History: PhotoModeler Software
Helps Knott Laboratory Reverse-engineer the Past
Reconstructionist
Vehicle Dynamic Characteristics of SUVs in On-Road, Untripped Rollover Accidents
Contribution of a Laterally Displaced Vehicle to the Post-Impact Deceleration of a Heavy Truck
Case Study: Follow-Up from Issue One
Case study - 2 in 3 out Case study Solution - 2 in 3
(c)2006CrashDataGroupInc
In real estate, it’s all about location, location, location. It would appear that, in publishing, it’s all about content, content, content. So, content has been our primary focus. The second issue of Collision is now bigger and better than the first with articles based on the 2007 ARC-CSI Conference in Las Vegas as well as articles offered by additional authors providing Collision’s readers with fresh, original content not found anywhere else.
Our motto is “out with old, in with the new.” We are not going to be satisfied with reprinting the same old thing; stale news releases and public records. We are going to continue to work to get new and original content to you in addition to meaningful, timely and useful testing data. Opinion pieces designed to generate discussion, test data you can actually use, topically relevant subject matter...that’s what we decided Collision would be all about.
Since the first issue, we’ve received input from a variety of readers on a number of topics. Some have asked some really good questions, and made some comments I thought I’d share:
“Your magazine Collision is a breath of fresh air. I have been reluctant to contribute to the [other magazine] because of their nonprofessional quality. I would like to contribute to your magazine on the topic of event data recorders. I wrote my Masters thesis on the court’s acceptance of event data recorder data.” On this note, we would love to see the reconstruction community get more involved with Collision. If you would like us to review and article or paper for inclusion in a future issue, please email it to editor@collisionpublishing.com.
Another correspondent wrote: “...You should expand to include engineer accident reconstructionists on a wider, more diverse editorial advisory board...” Our content advisory board members are volunteers from the larger reconstruction community and we welcome others who can contribute in a meaningful way. We don’t differentiate between one reconstructionist and another based on some perceived or self-serving description or a non-reconstruction specific background.
Another comment was: “Not too bad but not at a very high tehnical (sic) level. Since NAPARS is filled with police accident reconstructionists, you obviously have to severely limit the technical level or else those guys will get lost. I do understand why it cannot be more technical.”
After I picked my jaw back up off the floor, I wondered if this individual was reading Collision or some other publication. The “tehnical (sic)” level of the content of Collision is aimed at Crash Reconstructionists regardless of background. We welcome the submission of a “more technical nature” if someone would care to share what that actually means. In the meantime, we find that reconstructionists of a variety of backgrounds have found Collision a valuable resource and while we’re open to constructive input, it seems we’re heading the right way if we’ve been able to generate that kind of angst in just one issue.
It all got me to thinking, wait until they see THIS issue! That takes us back to content, with this issue of Collision you’ll find the digital file with some 2+ gigabytes of content including 70 crash video clips, over 1,000 digital photos, Vericom data, Stalker data, IST data, CDR data and PowerPoint presentations from the 2007 ARC-CSI Conference. You won’t find THAT anywhere else but Collision!
Scott B. Baker Managing Editor rashdatagroup.comAll advertisement are 4-color and professionally printed using a digital press.
With this issue of Collision, you will find a digital file with more than 2 gigabytes of content including 70 crash video clips, over 1,000 digital photos, Vericom data, Stalker data, IST crash data, CDR System re-trieved data and PowerPoint presentations from the 2006 ARC-CSI Crash Conference.
Scott Baker Publisher, Managing Editor sbaker@crashdatagroup.com
Sean Haight Editor
Tonya Baker Advertising Account Manager
W. R. Rusty Haight Collision Safety Institute
Brad Muir Collision Safety Institute
John Meserve President, NAPARS
Tom Szabo Biomechanical Reasearch & Testing
Judson Welcher Biomechanical Reasearch & Testing
All rights reserved, © 2006 Collision Publishing LLC. The opinions and conclusions expressed in this publication and attached data disk in articles attributed to specific authors are the opinions and conclusions of the authors noted and not necessarily the editorial staff or reviewers of those documents. Really, facts belong to everybody, any other opinions to us. The distinction is yours to draw...otherwise, the opinions expressed herein are not necessarily those of any employer, not necessarily ours and probably not necessary. Dissenting opinions, discussion or conclusions which may express an adverse position to those expressed herein by specifically cited authors can be addressed in writing by sending an email to the editor at editor@collisionpublishing.com or sending a “regular mail” letter to us at 118 Lake Street South Suite G, Kirkland, WA 98033. By sending an email to any or our email or snail mail addresses listed in this publication you are agreeing that: (1) we are by definition, “the intended recipient” (2) all information in the email is ours to do with as we see fit and make such financial profit, political mileage, or good joke as it lends itself to. (3) This overrides any disclaimer or statement of confidentiality that may be included on your original message. The entire physical universe, including this publication and attached data disk, may one day collapse back into an infinitesimally small space. Should another universe subsequently re-emerge, the existence of this publication and data disk in that universe cannot be guaranteed.
Past ARC-CSI Crash Conference CDs are available for purchase at www.crashdatagroup.com
Digital data is required for all ad submissions. Preferred format is EPS and PDF. Files should be press optimized, converted to CMYK, and have all fonts embedded. Publisher shall have no obligation or liability to Advertiser of any kind (including, without limitation, the obligation to offer Advertiser makegoods or any other form of compensation) if an as is supplied to Publisher by Advertiser in any format other than PDF. We will also accept AI, PSD, TIFF, and DOC file types. Please be sure to include a copy of all fonts used in your art work. Publisher will not supply a faxed or soft proof for Advertiser-supplied files. Advertiser is solely responsible for preflighting and proofing all advertisements prior to submission to Publisher. If Publisher detects an error before going to press, Publisher will make a reasonable effort to contact Advertiser to give Advertiser an opportunity to correct and resubmit Advertiser’s file before publication.
Accepted Media: Files may be submitted on a CD-ROM or FTP’d to the publishers FTP site. The FTP service requires a User Name and Password. Please contact Collision Publishing to get the log in information.
For complete and up-to-date information regarding advertising in Collision, please refer to our web site: www.collisionpublishing.com/advertising.html
Crash Data Group Inc
PO Box 892885
Temecula, CA 92589
Toll Free: 800-280-7940
Email: sbaker@crashdatagroup.com
Web: www.collisionmagazine.com
Advertise on the #1 web site for accident reconstruction and traffic accident investigation - The ARC Network. For complete information visit: www.thecrashhub.com/join
Summer has ended, so quickly it seems, but the good news is that with fall comes another exciting edition of COLLISION: The International Compendium for Crash Research, loaded with up-to-date news, research and crash test data as well as a disc containing hundreds of photographs and some 70 crash test videos. It’s the kind of information and material you can’t get anywhere else and that NAPARS is committed to delivering to its world-wide membership. I have received overwhelmingly positive responses from our members since the first edition hit their mailboxes in June and new membership applications are at a record high.
During these past two years, the NAPARS Board of Directors has supported my efforts to expand our organizations profile and presence in the accident reconstruction community nationally and world wide. We heeded the complaints that we were geographically limiting our efforts and not living up to the “National” in our name.
Our first on-the-road step was to join with the Ohio State Highway Patrol in conducting a successful three-day crash conference in Columbus in May 2005. A majority of the attendees had never had the opportunity to see live crash tests until Rusty Haight (left, showing on-lookers an air bag “burn”) and his CSI crew did their usual bang-up presentation. After participating with four other groups in our annual AR conference in September 2005, we went on the road again to Augusta, Maine and partnered with the Maine State Police to present a highly-praised 40-hour Commercial Vehicle Dynamics class this past May.
Maine State Police Colonel Craig Poulin welcomes yours truly to the agency’s headquarters (below, left). Participants were able to get “up close and personal” with several big rigs, and even got a chance to drive a tractor-trailer combination (below center, right).
Those were our first steps, and we want to take more and bigger steps in all directions. Where next?? If you have an idea to meet or supplement accident reconstruction education and training in your part of the country, let us know. We are willing to work with any agency or group toward the common goal of improving the levels of skill and knowledge in our profession. See you at a future NAPARS-sponsored event!!
John H. Meser ve PresidentFor more than 12 years,The CADZone has provided 1000’s of law enforcement officers with the highest quality software for creating crash and crime scene diagrams.
The new Crash Zone Version 8.0 is so much more than just diagramming! It features a Momentum Calculator,Skid Analysis,a realistic 3D body poser,and the ability to import measurements from a variety of sources,including total stations,Laser Technology devices,and iWitness photogrammetry software.Electronic,state accident reports with integrated Crash Zone diagrams are also available!
Greg South,Newtown P.D.(PA):
New calculation tools just for reconstructionists, including Skid Analysis,Combined Speed Calculator, and Momentum Calculator! (
“Always imitated, but never duplicated - your software keeps getting better.”
“Despite the other diagram software on the market,yours can’tbebeat!Thanks for keeping it simple and easy to use!”
ARC Network - www.accidentreconstruction.com - Founded in 1998, the ARC Network is the largest Internet portal web site for the industry of accident reconstruction and traffic accident investigation. This organization provides resources and information for all areas of accident reconstruction and traffic accident investigation including research, news, expert discussion, events, products and services, expert witness directory, book store, education services directory, police department directory and a member’s only section that provide valuable online databases and tools to assist in daily accident reconstruction requirements.
ASPACI - www.aspaci.org.au - ASPACI is a non-profit society whose members have joined to share in the challenge of analysing the complicated issues of road collision investigation. ASPACI was formed in 1991 by a group of police and collision reconstructionists who felt there was a need for information sharing between police and associated institutions in the Australia, South East Asia, New Zealand and South Pacific area. Its primary function is to assist in the spreading of knowledge and enhancing expertise in the science. The Association seeks to promote a professional approach to collision investigation and to encourage the highest standard of ethics, integrity and honesty among its members.
CA2RS - www.ca2rs.com - CA2RS was created to provide accident reconstruction training and resources to its members. There is a growng need for more local resources in the accident reconstruction field. This organization has grown to over 260 paid members and are always looking for ways to strengthen and increase their membership.
CATAIR - www.catair.net - The Canadian Association of Technical Accident Investigators and Reconstructionists (CATAIR) was founded in 1984 to meet the growing demand for a professional organization that subscribed to a code of professional conduct. CATAIR membership is open to individuals involved in all aspects of road safety. Current membership is in excess of 250. presently, our membership is comprised of active and retired police officers, professional engineers, automotive engineers, engineering technologists, private consultants and representatives from the insurance industry.
FARO - FARO, officially known as Forensic Accident Reconstructionists of Oregon, was chartered in June 1994 under the laws of the State of Oregon. Officially designated as a non-profit, mutual benefit corporation, FARO exists as a voluntary professional association of individuals who practice forensic reconstruction in Oregon.
IAARS - www.iaars.org - The International Association of Accident Reconstruction Specialists was started in 1980 as a training and networking resource for members. It is believed we are the oldest such crash reconstruction organization. Our membership consists of Law Enforcement as well as Engineers and PhD’s. We have members from 38 states as well as Canada, England, Singapore, Spain and West Australia. Our association is a little different from many others in that we have a peer review process for membership. This peer review is a specific area of judicial consideration when admitting an expert opinion in court. The annual seminars, held in different locations nationwide, not only provide instruction on specific topics by experts in the field, they provide face-to-face discussions on how other people approach reconstructing a crash. The old saying “two heads are better than one” is definitely true in accident reconstruction. The next seminar will be the Big Event in Houston, TX in September 2006.
IATAI - www.iatai.org - The Illinois Association of Technical Accident Investigators was born after several members of the Illinois State Police completed their training in traffic crash reconstruction. They saw the need for a way to share information, experiences and grow as Accident Investigators. These founding members set four goals for the organization; 1. To provide professional standards for members investigating accidents; 2. To promote the continued expansion of members knowledge in the area of technical accident investigation.; 3. To promote Traffic Safety through accident investigations; and 4. To promote improved capability of each agency through training and idea exchange. Since 1986, IATAI has held annual training conferences throughout the State of Illinois. These conferences serve to bring valuable training from across the country into Illinois. Our membership reaches out to 26 states in both the public and private sector.
MATAI - www.mdatai.org - MATAI continues to be innovative and willing to meet the needs of its membership. Over the years, MATAI has grown into a internationally renowned organization. MATAI has sponsored or co-sponsored numerous conferences that have been attended by fellow investigators from around the world. These conferences have included relevant and at times, spectacular test collisions involving cars, trucks, motorcycles, pedestrians, bicycles, and even a Metrobus.
MATAI - www.matai.org - Founded in 1988, the Midwest Association of Technical Accident Investigators (MATAI) was formed to provide a professional affiliation for individuals who have a primary interest in the technical aspects of motor vehicle traffic collisions. The Association is dedicated to the exchange of information and ideas to improve investigative techniques and procedures In addition, MATAI provides a communication source between other individuals and affiliations involved in this vital area of public safety. MATAl’s primary objective is to meet the needs of the people directly involved in the initial investigative process.
NAPARS - www.napars.org - The National Association of Professional Accident Reconstruction Specialists is a non-profit organization whose members have joined together to share the challenge of dealing with the complex problems of accident reconstruction and to upgrade and ultimately professionalize the accident reconstruction field. NAPARS is open to all persons who are interested in the fields of traffic accident reconstruction and highway transportation safety. Present membership exceeds 1,100 and includes police officers, engineers, consultants and government safety personnel. NAPARS is also the Premier Industry Partner for Collision Magazine.
NATARI - www.natari.org - The National Association of Traffic Accident Reconstructionists and Investigators (NATARI) was established as a non-profit Pennsylvania corporation in 1984. The goal of NATARI is to provide a source of information to be shared among accident professionals on a national basis. This goal is reflected in the NATARI logo comprised of an outline of the nation, a broken wheel signifying a traffic collision and the words “to solve for safety”.
NJAAR - www.njaar.org - The New Jersey Association of Accident Reconstructionists started in 1991 as a dream of a few New Jersey police officers who wanted to create a network of individuals with interests in the field of accident reconstruction. Its membership has grown to encompass individuals from many states in all fields of accident reconstruction. We have a large membership base that includes many professionals not only from the police community, but the private sector as well.
NYSTARS - www.accidentreconstruction.com/nystars - The mission of NYSTARS is to organize into one body, professionals in the field of ground vehicle collision reconstruction and related activities, who support and maintain a high standard of ethics, integrity, credibility and honor in the field of collision reconstruction; to encourage training programs relating to motor vehicle collision investigation and reconstruction through research and communication of matters of mutual interest; to promote traffic safety and foster a spirit of brotherhood among the organization’s members.
SATAI - www.satai.com - Established as a non-profit corporation in 1982, the Southwestern Association of Technical Accident Investigators, Inc. goal is to encourage and promulgate the development of professional technical accident investigation and highway safety; promote research and development of programs leading to better technical accident investigations; and to promote the development and dissemination of new knowledge in the fields of traffic safety, accident investigation and reconstruction. The Association is organized and operated for scientific and educational purposes.
SOAR - www.accidentreconstruction.com/soar - An organization of accident reconstructionists dedicated to increasing knowledge, enhancing auto engineering, improving traffic safety and the free exchange of information. Membership is open to all people in all aspects of traffic transportation. Membership is comprised of active and retired police officers, professional engineers, automotive engineers and private consultants. We have members throughout the world.
TAARS - www.accidentreconstruction.com/taars - TAARS was organized in 1986 as an organization dedicated to advancing Research, Knowledge, Education and Safety in the field of accident reconstruction. TAARS is a broad based organization with members from many different fields, including law enforcement, private investigators, engineers and attorneys. TAARS has a current membership of over 220 members.
WATAI - www.accidentreconstruction.com/watai - WATAI was formed in 1979 and is the oldest of the technical accident investigation organizations in the United States. Twenty-five plus years since its inception WATAI remains true to its original mission; “to promote education and research in the field of accident investigation; to encourage interdisciplinary communication between the practitioners of accident investigation; and to cooperate and participate with other organizations dedicated to the science of accident investigations”. To that end WATAI sponsors semi-annual conferences. The Spring Conference usually takes place in May and the Fall Conference generally in the onth of October.
(c)2006CrashDataGroupInc
The forensic scientist is constantly faced with comments like, “that is not what the witness testified to” or “why should we believe you and not the witness, after all, they were there”.
The problem has always been that the forensic scientist offers information based strictly upon the facts they have been successful in locating, then determining how they best fit together. The witness on the other hand is offering testimony based upon what he remembers seeing, hearing, feeling, smelling, thinking, etc. As we all know, the senses are often fooled by very simple things, like color changes caused by the type of gas in a light bulb altering colors, or their own self-darkening eyeglasses. This indicates that the eyewitness makes for a less than credible witness. But is there more to this problem?
For the moment, let us ignore that very real problem of the witness that is less than truthful for a variety of reasons, and focus only on those that are doing their utmost best to aid the investigation by providing the complete and truthful circumstances, as they recall them. My concern is primarily
with traffic related incidents. However, this discussion pertains to any eyewitness account.
As with any collision investigator and reconstruction professional, I base my opinions and conclusions on the factual evidence. If I can reconcile the statements of the involved parties and witnesses, so much the better. However, a complete reconciliation seldom occurs.
Needless to say, the lawyers are always excited and concerned when they find out that your findings differ from the statements of their client or their opposition. Especially when they completely contradict on some issue. I have always attempted to explain this problem in a manner that the attorney and the court could understand. This explanation was based upon my 30 years of experience doing this job. While attending college to get my next degree, I was forced to take a Developmental Psychology course. To say that I was less than enthused by the prospect would be a definite understatement. However, this course has resulted in some interesting background information. I pass this along to you, to use as you see fit, I hope it aids you when attempting to explain this same situation.
(c)2006CrashDataGroupInc
One of the reasons that the witness and involved party statements don’t exactly match the factual account of the incident is due to what Jean Piaget referred to as Cognitive Equilibrium or Equilibration. This theory states that “all intellectual activity is undertaken with one goal in mind: to produce a balanced, or harmonious, relationship between one’s thought process and the environment” (Schaffer, 2002). This ability is based upon the constructivist theory that states “an organism acts on novel objects and events and thereby gains some understanding of their essential features” (Schaffer, 2002).
What this means is that for a person to be normal, the mind constantly strives to place everything it is aware of, into a format that is logical to it, based upon the person’s education and experiences. Therefore, a person will “remember” the incident based upon their experiences rather than based strictly upon actual events. When presented with the actual events, based upon facts; and the reason for this chain of events is offered to the person, they will frequently remember the events in the proper order. However, “scientific reasoning is a late developing ability that even many adults fail to master” (Schaffer, 2002). Because of this, even when presented with the facts and the reasons you know those facts to be so, some will be unable to understand and “remember” the events properly.
Stated another way, it is the Kantian point of view that causality is based upon selective interpretation of external events as filtered through the perceiver’s framework of logical reasoning. In 1972 Trankell studied the post-hoc fallacy and “suggested that our causal judgments are inevitably based upon a personal interpretation of data because the logical completion mechanism that fills in causal gaps are based on patterns of earlier experiences” (Walton, 2005, Pg. 214). This is simply a description of the post-hoc fallacy, in offering the mistaken notion that simply because one thing happens after another, the first event was a cause of the second event. Post hoc reasoning is the basis for many superstitions and erroneous beliefs. (An example is that after it rains, you determine it was caused by you washing your car). This is exactly what the eyewitness is doing unintentionally. So that their mind can file away the incident, it must first edit the event, so that it is logical to them. This requires them to make leaps of judgment as to why things occurred. Therein lies the origin of the false testimony.
Knowing this, the best thing you can do to assist your case, is to talk with the witness or involved person and obtain a complete statement for their perceived chain of events. It will also be helpful in court later, if you obtain an educational history of the witness, and if possible, a similar experience history. From this information, you can make an intelligent case for your opinions and the differences from the witness’s recollections. You may even be able to offer credible reasons for the witness’s errors in perception, based upon their education and experiences.
Vygotski explained that the reason that the witness can reassess their opinions is because of what he termed “Tools of Intellectual Adaptation” and “Guided Participation”. These “tools” permit a person to alter their perceptions based upon newly acquired knowledge. The “participation” permits the person to absorb the information presented by a person perceived as being more knowledgeable.
Jean Piaget was a Swiss scholar that lived from 1896 until 1980 and began studying intellectual development in the 1920’s. Lev Vygotski was a Russian developmentalist and lived from 1896 through 1934 and was Piaget’s most outspoken critic. The two researchers have combined to offer much insight into the cognitive development of the human brain. Immanuel Kant was a German Philosopher from Konigsberg (now Kaliningrad) in East Prussia and lived from 1724 until 1804. Kant is often considered one of the greatest, and is one of the most influential, thinkers of modern Europe and the last major philosopher of the Enlightenment.
REFERENCES:
Douglas N. Walton (2005). Informal Logic – A handbook for Critical Argumentation. Cambridge University Press Schaffer, D. (2002). Developmental Psychology. (Sixth Edition), Wadsworth Group, Thompson Learning Inc.
-George Ripsom
is a forum for opinion and commentary. The opinions expressed may not reflect a widely accepted or adopted view or position but may be meaningful to the author and many in the community. In any case, they are offered to generate discussion and or debate within the community. Your responses or submissions for this periodically recurring feature are invited.
Calculating the Principle Direction of Force “pdof” relative to a vehicle involved in a collision, is important for a number of reasons. Not the least of these is demonstrating that the analysis has complied with Newton’s Three Laws of Motion while providing for a basis of an occupant motion or kinematics study.
Arguably the most commonly used formula used to calculate the pdof is the following:
ist has met the test of Newton’s Third Law of Motion. That is, the force applied to each of the vehicles was equal and specifically opposite when the vehicles are positioned along their respective approach headings at the point of maximum engagement.
Where:
v’ = the vehicle’s post impact velocity
Delta Theta = the change in direction between the vehicle’s approach and departure angles.
Delta v = the change in velocity the vehicle experienced as a result of the collision.
In the confines of collision reconstruction, the pdof is used to describe the direction of the force that was applied to the vehicle during the collision. The pdof is measured relative to the direction the center mass was traveling on impact, typically along the longitudinal axis of the vehicle for which the pdof is being determined. When using this convention, the longitudinal axis (running front-to-rear) is such that an angle of 0° is at the front of the vehicle while the angle of 180° is to the rear. When the pdof angle is expressed as a positive number it identifies a pdof that intersects the longitudinal axis from the right, while a pdof expressed as a negative number identifies an angle intersecting the longitudinal axis from the left. However, it is import to note, without additional analysis, there is no significance whether the angle returned by the pdof equation is either positive or negative.
The pdof is used, from a purely mathematical or physics based perspective to demonstrate that the reconstruction-
From an occupant motion perspective, the pdof is used to analyze the kinematics or motion of occupants in a vehicle during the collision as a result the collision forces. Relative to the vehicle, the occupants in a car will appear to move in a line opposite to the force applied, or toward the pdof. For example, a force applied to the right front quarter of a vehicle will result in the appearance of the occupants moving inside the car, to the right front. Therefore, an unrestrained right from passenger might be expected to make contact with the right “A” pillar while an unrestrained driver might be expected to move toward the review mirror mounted on the windshield.
MATHEMATICAL BASIS: The most common pdof equation is simply an application of the “Law of Sines” where the values for two sides of a triangle and one of the angles are known. In its application to collision reconstruction, there are situations where the equation will return an incorrect answer. This occurs when the actual pdof exceeds 90°. To better understand why this occurs we need to first understand where the Law of Sines and the pdof equation comes from.
LAW OF SINES PROOF:
To understand where the Law of Sines comes from we simply start with a triangle where the three sides have been labeled a, b and c, and the angles have been labeled A, B and C. Then divide the triangle into two right triangles by adding h.
Having created two right triangles we know the following
Solve for h:
Since h equals both b × sin A and a × sin B then b × sin A = a × sin B
Divide both sides of the equation by a×b
Leaving the first equality of the Law of Sines.
Redefine the altitude (h)
Solve For h:
With h redefined, the same relationship that was established with angles A and B can now be established with angles B and C. Furthermore, if the triangle is rotated so h is once again vertical as in it can easily be seen that the same right triangles exists as before. Since h equals both c × sin B and b × sin C then c × sin B = b × sin C
Using the following example, one can demonstrate how the pdof formula is derived from the Law of Sines. Vehicle #1, which weighs
3,500 lbs (1587.6 kg) is traveling due east along an impact heading of 0°. It is struck by Vehicle #2, which weighs 4,000 lbs (1814.4 kg) and is traveling north with an impact heading of 90°. Vehicle #1’s
post impact speed is determined to be 25 mph (40.2 km/h) while its post impact heading is 35°. Vehicle #2’s post impact speed is determined to be 30 mph (48.3 km/h) with a heading of 70°.
Using a mathematical approach based on conservation of momentum and the equations for determining delta v and the pdof, we find the following:
Vehicle #1 Vehicle #2 Vehicle #1 Vehicle #2
Speed 32.2 mph 40.7 mph 51.8 km/h 65.6 km/h
delta v 18.5 mph 16.2 mph 29.8 km/h 26.1 km/h
pdof 50.7 ° 39.3 ° 50.7 ° 39.3 °
Table 1
The pdof angles reported in table 1 suggest that the quantity have a positive value. In fact, within the guidelines of the conventions for measuring angles relative to motor vehicles as found in SAE J1594, there is a distinction between a positive and negative angle measured relative to each vehicle. However, as mentioned previously, there is no significance whether the result of the pdof equation is either positive or negative without further analysis. In this particular example the results of the pdof equation are both positive, yet the pdof for Vehicle #2 should be recorded as a negative value.
The solutions reached mathematically may also by analyzed using a graphical approach using a vector diagram solution, as seen in . Using this approach, we can see where the change in momentum (delta P) is found on the angle established by the pdof when measured from the vehicle’s impact vector.
To better understand the relationship of the different vectors for a single vehicle, the portion of the vector diagram that applies only to Vehicle #1 is isolated.
The sides of the triangle may be labeled a, b and c, and the angles A, B, and C, just as the triangle was that was used to describe the Law of Sines.
Having established the triangular relationship between Vehicle #1’s pre and post impact vectors and the P vector, the Law of Sines can be used to solve for the pdof or in this case angle B. Starting with the first equality of the Law of Sines, you can derive:
The values of previous equation may now be replaced with the values we know from our problem. Referring back, the following may be concluded:
Angle B is equivalent to the pdof
•Side b is equivalent to the post impact momentum vector
•Side a is equivalent to the P vector
Angle A is the difference between Vehicle #1’s pre and post impact directions. Therefore, since Vehicle #1’s pre-impact heading was 0° and its post impact heading was 35°, the value of A equals 35°.
Additionally, since we are essentially dealing with a ratio between sides b and a, the units used in the equation, i.e. mph, km/h, fps, etc., is unimportant, providing that the units remain consistent throughout the equation.
Having identified the origins of the pdof equation based on the Law of Sines the situation where it will produce an incorrect answer can be examined.
Vehicle #1, which weighs 3,500 lbs (1587.6 kg) is traveling due east along an impact heading of 0°. It is struck by Vehicle #2, which weighs 4,000 lbs (1814.4 kg) and is traveling northeast with an impact heading of 40°. Vehicle #1’s post impact speed is determined to be 25 mph (48.3 km/h) while its post impact heading is 35°. Vehicle #2’s post impact speed is determined to be 30 mph (48.3 km/h) with a heading of 20°.
As before, using a mathematical approach based on conservation of momentum and the equations for determining delta v and the pdof, we find the following:
However, when we attempt to define the altitude and spilt the triangle into two right triangles by adding h, the altitude, falls completely outside of the triangle.
As expected, the results of the graphical solution to the problem are nearly identical to the mathematical results. However, there is one distinct difference between the mathematical solution in table 4 and the graphical solution in. The mathematical results yielded an answer of 82.5° for the pdof of vehicle #1 while the graphical solution produced an answer of 97.5°, a 15° difference. When looking at the vector diagram it can seen that the delta P’s for both vehicles are equal and opposite. Therefore the pdof value produced by the graphical solution has to be the correct value, demonstrating the problem associated with the pdof formula based on the Law of Sines.
Using the same approach used in the first problem, the portion of the vector diagram relating to vehicle #1 is isolated, and the sides relabeled a, b, and c, and the angles A, B, and C.
Yet, when known values are inserted into the Law of Sines equation, it still appears to be valid.
A = departure angle … 35°
B = pdof from the vector diagram … 97.5°
C = 180 – (A + B) … 47.5°
a = delta v … 17.4 mph
b = post impact speed … 30 mph
c = impact speed … 22.3 mph
However, when the pdof is calculated mathematically, the incorrect answer is produced.
The reason this occurs is that when measuring from the X axis, there are two angles that have a sine of .9914. The first is the reference angle, which is the acute angle formed by the terminal side and the X axis, while the second angle is the obtuse angle formed by the terminal side and the X axis. Both of these angles will have the same sine value. Yet when the inverse sine is taken of that value it will only produce a single result, which will always be less than or equal to 90°.
The relationship that exist between the two angles measured off either side of the terminal side as illustrated. brings us back to the derivation of the Law of Sines and the Ambiguous Case.
In the example the pdof equation solved for angle B in triangle made up by the vectors associated with Vehicle #1. Therefore, since side a is greater than h, there are two possible triangles that can be made by angle A, side b and side a. The first consists of sides a, b, c and angle B, while the second is made up of sides a’, b, c + c’ and angle B’.
As shown, the value of B is greater than 90° (97.5°) while B’ is less than 90° (82.5°). Having already established that an angle and its reference angle have the same sine value and the inverse sine of that value will only produce a result less than 90°. Then in our example, the pdof equation will return the value for B’ as opposed to the value of B, the angle the pdof equation was intended to solve for. The correct angle can be determined simply by subtracting the angle produced by the pdof equation (the value of B’) from 180°. This “error” occurs because the arcsine or inverse sine function is not a true inverse function. A true inverse function, such as addition and subtraction will have a one to one relationship. For a function to have a one to one relation any given set of values in conjunction with the function will only produce a single result, for example, 3+2 = 5 and 5 – 3= 2.
To determine if a function has a one to one relationship it has to pass the “horizontal line test.” In order for a function to pass the horizontal line test, when the function is graphed, a horizontal line drawn across the plot may only intersect with the function in one location. In the graph to the left the functions x² and x³ have been plotted along with a horizontal line. In the x² plot the horizontal line intersects with the plot in more than one location. Whereas, in the x³ plot, the horizontal line only intersects with the graph in a single location. Hence the function x³ has a true one to one relationship. 2³ = 8 while -2² = -8, and Cubed root of 8=2 and cubed root of -8=--2. Whereas the function x² does not have a one to one relationship since both -2² and 2² = 4, while there is no real solution to square root of -4.
When the sine function is graphed it is easily seen that does not pass the horizontal line test. In fact there are an infinite number of angles that will produce any particular sine value. For instance, the sine of 30° equal ½ as does the sine of 150°, 390°, 510°, 750°, 870°...∞.
Since the sine function lacks a one to one relationship, there cannot be an Inverse function, yet one exists. This is possible since the inverse sine function is based on a sine function whose values have been limited between ½×-pi (negative 90 degrees) and ½×-pi (positive 90 degrees). When the sine function is limited in this manner it passes the horizontal line test. However, since the inverse sine function is based on the limited sine function, it is only capable of returning a value between -90° and 90.
Logically the next question that arises is: How is the error associated with the pdof equation easily identified? Since the limitations associated with the Law of Sines based pdof equation only occurs when the pdof of a vehicle exceeds 90° it is not a common occurrence. Nonetheless, it is a real problem that will produce an incorrect value unless it is identified and taken into account during a proper analysis.
There are two conditions that have to exist before the limitations associated with the Law of Sines based pdof equation needs to be addressed.
•The first is that the approach headings of the vehicles must be within 90° of each other.
•Secondly, the vehicle’s post impact velocity for which the pdof is being calculated must exceed its impact velocity.
If both of these conditions are not present the pdof cannot exceed 90° and the possible error does not need to be considered.
Returning back to the example, since the approach angles of the two vehicles are within 90° of each other and the calculations show the post impact speed of Vehicle #1 exceeded its impact speed, then the possibility exists there is an error with the pdof calculations. Thus raising the question: How do we know if our pdof calculation is wrong? As in the example.
There are two similar approaches that can be used, based on information already known, to determine if the calculated pdof will be incorrect.
To illustrate both methods the portion of the vector diagram associated with vehicle #1 is once again isolated.
The simplest way to avoid having to address the possible “error” associate with the Law of Sines based pdof equation is to use a different equation. Fortunately this may be done by using a pdof equation which is based on the Law of Cosine.
c² = a² + b² - 2 × a× b × cos(C)
As before begin with a triangle with the sides labeled a, b, and c and the angles labeled A, B, and C. However, unlike the with the Law of Sines, this triangle may take on any shape as illustrated.
Side a in either of the above triangles will be comprised of the sums of side a1 and a2. In both of the above triangles, side a1 will represent the side of the right triangle that is formed by angle B, side c and h, whereas a2 will be defined by angle C side b and h. Both a1 and a2 will represent a positive value, whose sums will equal the length of side a. a1 will represent a positive value whose length is greater than the length of side a, while a2 will represent a negative value (the cosine of an angle between 90° and 270° will produce a negative value). Once again when the values of a1 and a2 are added together, the result will equal side a.
The length of side a may be expressed as:
a = c× cos(B) + b× cos (C)
Multiply both sides by a
a² = a×c× cos(B) + a×b× cos (C)
Complete the process for sides b and c
b² = b×a× cos(C) + b×c× cos (A)
c² = c×a× cos(B) + c×b× cos (A)
Add a² + b²
a²+ b²= a×c× cos(B) + a×b× cos (C) + b×a× cos(C) + b×c× cos (A)
Simplify
a²+ b²= c²+2×a×b× cos (C)
Subtract 2×a×b× cos (C) and rearrange to solve for c²
c²= a²+ b²-2×a×b× cos (C)
When using the vector diagram for Vehicle #1 in both the previous examples, the Law of Cosines may be derived to solve for the pdof of Angle B.
Starting with the variation of the Law of Cosines written to solve for the length of side B.
b²= a²+ c²-2×a×c× cos (B)
Add 2×a×c× cos (B) to both sides of the equation
b²+2×a×c× cos (B)= a²+ c²
Subtract b² from both sides
2×a×c× cos (B)= a²+ c²-b²
Divide both sides by 2×a×c
Take the Inverse Cosine of both sides to solve for B, substitute known values
Finally, like the Inverse Sine function the Inverse Cosine function is based on limited or restricted values of the Cosine function.
Nonetheless, unlike the Inverse Since function which only returns a value between -90° to 90°, the Inverse Cosine function returns a value between 0° and 180°. Hence the problem of producing the wrong result is not an issue with the Inverse Cosine function since the pdof, cannot exceed 180°. However, since the Inverse Cosine function will only yield a positive result, the determination as to the proper signage of the calculated pdof still has to be determined, which will be addressed at a later time.
NOTES:
1 In the example, the approach heading of Vehicle #2 and the departure headings of both vehicles are measured relative to an arbitrarily established base line which is also the approach heading for Vehicle #1. In this case the angles were measured in a counterclockwise direction from the established east 0° heading
2 In the example, the approach heading of Vehicle #2 and the departure headings of both vehicles are measured relative to an arbitrarily established base line which is also the approach heading for Vehicle #1. In this case the angles were measured in a counterclockwise direction from the established east 0° heading.
(c)2006CrashDataGroupInc
Timothy J. Reust Accident Science, Inc.INTRODUCTION:
Modern passenger vehicles with ABS brakes are capable of generating up to a 1.0g deceleration rate (the range in general is 0.8 to 0.9g) on a dry road surface that is in good condition. This paper outlines recent testing with a variety of passenger vehicles that include sedans, minivan, SUV and pickups. The tests were conducted on dry road surfaces that were in good repair. The deceleration rates of the vehicles were tested during straight line braking and also during yaw events.
INSTRUMENTATION:
The instrumentation used for this paper consisted of two different systems.The Racelogic VBox III which is a GPS based device that measured time, distance and speed and has a sample rate of 100 Hz. Used during some of the tests with the VBox was an add-on unit called the IMU (inertial measuring unit) which provided 3-axis acceleration and yaw, pitch and roll data. Some of the early testing was done using a Datron V1 Optical 5th wheel that measured time, distance and speed. This is a 2axis system that measures in both the x and y direction and was setup record at a rate of 100 Hz. A tape switch was placed on the brake pedal and connected to the systems to record a brake application signal.
Instrumentation that continuously measures time, distance and speed provides data that can be used to calculate the deceleration during different phases of a skid test. A device that can accurately measure speed when
traveling in a curve or in a yaw provides data that can be used to calculate the deceleration during these types of events. Both of the systems outlined above worked well for the yaw testing.
During the straight line brake tests, the vehicle was accelerated up to the desired speed and then the brakes were applied firmly and held until the vehicle came to a stop. The yaw tests were done in two different ways. The first method was to bring the vehicle up to speed then introduce rapid steer in one direction until the vehicle began to yaw, the steering was held in the same position till the end of the test. The second method was to bring the vehicle up to speed, introduce rapid steer in one direction and then fully apply the brakes and continue to brake and hold the steering in the same position.
(c)2006CrashDataGroupInc
The deceleration rates for the straight line braking were calculated by using a speed and time sample at the beginning of the steady state speed decrease and then using a speed and time sample at the end of the steady state speed decrease. The “instantaneous” deceleration ranges for the yaw tests were calculated using speed and time with an offset of 0.5 seconds. This method used a group of five continuous speed readings that were averaged then a second group of five continuous speed samples that were averaged with the time offset between the samples as 0.5 seconds.
This method was done along the entire data string. An average deceleration rate during the first 100 feet of the yaw was calculated by using the speed at the beginning of the yaw and the speed after 100 feet. The criterion for the beginning of the yaw was when the yaw rate initially peaked and/or the lateral acceleration reached a maximum steady state.
Many papers have been written on the topic of deceleration rates and many others will follow. It is the hope of this author that future work will continue to add the database and that the future papers will outline the equipment that was used in measuring and the methods used in deriving the deceleration. This paper has outlined that deceleration rates for the straight line brake tests with ABS were primarily well above the 0.7g range. This value should generally be expected with modern vehicles on a dry road surface. Very little has been written on the topic of deceleration rates during a yaw event. As can be seen from the results in this
paper the “instantaneous” deceleration rate tends to fluctuate during a yaw event. Additional testing needs to be conducted in this area to help identify trends and values.
THe author would like to thank James M. Morgan for his help with the vehicle testing and the data analysis.
Graphs from the yaw tests have been included for the readers review. Most of the graphs include vehicle speed, yaw rate, lateral acceleration and the calculated deceleration rate. The following tables outline the vehicles that were tested, the type of test and the resulting deceleration ranges and rates.
(c)2006CrashDataGroupInc
Sport Utility Vehicles Non-ABS
Vehicle Yaw Decel Range (g)
2005 Chevrolet Suburban steer .19 to .32
2005 GMC Yukon XL steer .1 to .35
2005 GMC Yukon XL steer .08 to .37
2005 Cadillac Deville steer .2 to .5
2006 Cadillac SRX steer .04 to .24
2005 Pontiac Bonneville steer .05 to .2
Vehicle Yaw 100ft Decel Rate (g)
2005 Chevrolet Suburban steer 0.283
2005 GMC Yukon XL steer 0.292
2005 GMC Yukon XL steer 0.303
2005 Cadillac Deville steer 0.383
2006 Cadillac SRX steer 0.194
2005 Pontiac Bonneville steer 0.16
Vehicle Yaw Decel Range (g)
2005 Pontiac Bonneville steer and brake .24 to .7
2006 Cadillac SRX steer and brake .16 to .67
2005 Cadillac Deville steer and brake .37 to .70 2005 GMC Yukon XL steer and brake .31 to .47
Vehicle Yaw 100ft Decel Rate (g) 2005 Pontiac Bonneville steer and brake 0.416
Cadillac SRX steer and brake 0.41
Cadillac Deville steer and brake 0.494
GMC Yukon XL steer and brake 0.394
PROBLEM:
During March of 2003, the author responded to a report of a single vehicle collision that resulted in the death of both occupants. Tire marks were observed on the pavement surface, leading to the position of the vehicle at final rest, that were unlike any catalogued or previously observed at the scene of a collision. The tire marks most closely resembled scalloped marks, typically produced by a deflated or under-inflated tire. When the tire pressures were checked on the overturned pick-up truck and found to be fully inflated, an explanation for the tire marks was sought. This paper seeks to explain the vehicle/operator action and mechanical equipment that produced the unique marks and their bearing on efforts to reconstruct the event.
BACKGROUND:
Text books, written for the field of motor vehicle collision investigation and reconstruction, catalog and identify numerous types of tire marks which are commonly found on pavement and ground surfaces after a collision event. Tire marks with scalloped characteristics are identified in many texts as having been produced by a flat or severely underinflated tire.
DYNAMICS:
The tire marks present at the scene indicated the vehicle was in a sideslip action, rotating clockwise, as it traveled down the road. Although crossover tire marks were observed, they could not be categorized as textbook yaw marks, as evidence of hard braking was apparent. The front tires of the vehicle produced curved marks on the blacktop surface that were consistent with ABS scuff marks, with some parallel striations. The
rear tires of the vehicle produced the unusual scalloped marks. As a result of the vehicle rotation and the transfer of weight, the left side tire marks were more pronounced than that of the right side tires. The left rear tire mark contained both parallel and diagonal sections of striations. Collision investigators are familiar with the known, that dynamic friction exceeds static friction. In this case, this relates to the tires in motion (rotating) as they slide along the pavement, as opposed to the tires being locked as they slide along the pavement. It appeared that those portions of the marks that contained parallel striations also exhibited a sharp curve in the mark, indicating an area of greater vehicle rotation. It also appeared that the portions of the marks that contained diagonal striations coincided with the rear tires traveling a path closer to that of the vehicle’s center of mass. The pattern described above repeated itself in the marks for approximately 145 feet, where the vehicle exited the roadway and overturned. Refer to figure 1 below, which is a photograph of the marks observed.
The vehicle involved in the collision was a 1998 Chevrolet 1500 series pick-up truck, equipped with a selectable 4X4 drive train. The bed of the pick-up was essentially un laden and the design of the truck results in significantly less weight on the rear axle. The transfer of weight during braking or maneuvers has an increased effect on this style of vehicle. The antilock brake system (ABS) was controlled by a Kelsey-Hayes manufactured brake pressure modulator valve (BPMV) assembly. This particular system has threechannels of control, with a speed sensor at each front wheel and a single speed sensor for controlling both rear wheels. In use, this means if one rear wheel locks, hydraulic pressure to both rear brakes will be reduced by the system.
The marks at the scene indicated that the individual wheel speed sensors at the front of the vehicle operated more rapidly and efficiently than the single speed sensor for both rear wheels. In effect, the front tires more closely followed the steered path of the vehicle, while the rear tires were subjected to the effects of varying friction as a result of intermittent wheel lock up and hydraulic pressure release. The left rear tire alternated from locked to rotating approximately every 5-10 feet. It is likely that this unusual action was felt by the operator, however, it is not likely to have affected vehicle control. The evidence at the scene indicated that this anomaly did not occur until after “crossover” and at a point where the sideslip was unrecoverable.
Regardless of the effect on vehicle control, having an understanding of the mechanical components helped to explain the operator action that produced the unique tire marks. This collision occurred on a relatively straight section of roadway, and for some unknown reason, the operator made steering inputs to the left and then overcorrected when steering back to the right. It was also apparent that the operator applied the brake pedal after steering right. The evidence indicated that the brake pedal remained depressed to the point where the vehicle exited the roadway.
In the collision at hand, calculations were undertaken to determine the speed of the vehicle. The speed was estimated utilizing two differentmethods. The first method utilized the slide to stop formula for the distance traveled on the pavement, where the marks indicated that the brakes were applied. When considering the speed loss prior to braking and speed loss as a result of running off the roadway and overturning, it was comparable with the
results of the critical curve speed formula. In both formulas a coefficient of friction was applied as a result of skid tests conducted at the scene. It was determined that the vehicle was traveling in excess of 70 MPH. In this investigation, the results of both methods were consistent with the circumstances of the collision. This bolstered previous studies that found the critical curve speed formula is valid when there is evidence of braking. It is unknown what effect, if any, the observed marks would have on a speed analysis in other cases.
TESTING:
An exemplar vehicle was obtained for the purpose of attempting to duplicate the marks that were observed at the collision scene. A similarly equipped 2000 GMC pick-up truck with the identical Kelsey-Hayes ABS and BPMV was used on a nearby airport runway. In a series of test runs the exemplar vehicle was driven to speeds ranging between 30 - 51 MPH and oversteer inputs were made to produce a sideslip. Once the vehicle began to yaw the brakes were applied with full force. The author was able to sense some undulation in the movement of the rear-end and two of the higher speed tests produced some marks that exhibited a very slight waver. However, the waver was short in length, due to the fact that each time the brakes were applied the motion of the vehicle was corrected. Specifically, the ABS would bring the vehicle under control and correct the tracking, therefore the yaw could not be sustained. Due to speed limitations and safety considerations the tests were unable to reproduce the vehicle action and the marks that were at the collision scene. The waver in the test marks, at lower speeds, was so slight that
they were not detectable in a photograph.
Although similar to scalloped marks produced by under inflated or flat tires, the pulsing action of the ABS was evident in the tire marks. The front tire marks alternated from darker to lighter in appearance, but continued in a relatively constant curved path, consistent with the rotation and travel of the vehicle. The pulsing of the ABS was also apparent in the rear tire marks as previously described. This produced a serpentine like mark that indicated the rear of the truck was undulating back and forth during the sideslip. The mark was a result of constant hard braking, weight shift and oversteer by the operator. The individual wheel speed sensors on the front of the vehicle apparently managed the ABS better than the single speed sensor for the rear wheels. It does not appear that this equipment contributed to the collision, as the marks were produced after a loss of vehicle control. In this investigation, the marks did not appear to have a significant effect on an analysis of vehicle speed. These marks had not been previously encountered, nor have they been encountered since. The actions, conditions and equipment that produced the unique marks have been explained. However, the testing was unable to duplicate the unique marks. A more conclusive finding of the effect on vehicle control or speed analysis would require an investigation of a number of collisions with similar roadway markings or additional testing.
“The evidence at the scene indicated that this anomaly did not occur until after “crossover” and at a point where the sideslip was unrecoverable.”
ABSTRACT:
In the analysis of motor vehicle collisions, involving passenger cars and/or light trucks, the determination of a deceleration rate is relatively simple to determine. Vehicles with conventional braking systems usually “lock up the tires” in pre-impact skids and the tires are often locked by vehicle damage post-collision. The percentage of the vehicle’s weight remains fairly constant during the pre and post-impact phases of the collision. Vehicles with ABS systems perform rather uniformly during “skidding” or hard braking maneuvers, and usually react predictably and similar to vehicles with conventional braking systems post-impact, assuming the vehicles are uncontrolled post-impact.
However, motorcycles have unique braking systems. There are two separate brake controls on the vast majority of motorcycles, and sometimes different types of brakes on each wheel. Some motorcycles have integrated braking systems and a few models have ABS systems. Previous studies have shown that it is difficult for the motorcycle operator to efficiently control two independent brakes during hard braking maneuvers. Add to that the reduced fine motor skills during stressful situations, such as impact avoidance, and it is even more difficult to effectively utilize the brakes on a motorcycle. Limited rider training in the proper use of motorcycle brakes further reduces the effectiveness of the vehicle/rider braking capability.
The author attempted to determine the most significant factors involved in the determination of the deceleration rate to be assigned to the motorcycle in pre-impact braking.
There have been significant advances in the design of passenger car and light truck braking systems. ABS systems are becoming more the norm than the exception. Motorcycle braking systems have been improved in the area of friction capability, but the overall braking system has remained relatively the same since the vehicle’s original design – a hand lever controls the front brake and a foot pedal controls the rear brake. In integrated systems there is some pressure applied to the front brake during activation of the rear brake and in some there is rear brake application during front lever actuation.
Motorcycle operator training is limited. Most motorcycle training courses are conducted at low speeds, usually under 35 mph, and at isolated locations, free from other traffic. The riders are taught how to control the vehicle’s clutch, throttle, and gear shift to start from a stopped position, and to accelerate from a stopped position. They perform low speed turns through a marked course. There is training in proper clutch and brake control to bring the vehicle to a safe stop during normal operating conditions. Few motorcycle operators receive any additional training before taking their vehicles on the road.
During the investigation of motor vehicle collisions, the investigator commonly attempts to quantify speed (energy) losses of the involved vehicles. For passenger cars and light trucks, the determination of an acceleration rate (positive or negative) is typically easier to determine than that of a motorcycle.
To stop a passenger car or light truck, the operator simply applies the brakes to their fullest. The vehicle typically will not roll over in hard braking maneuvers and the operator can release the brakes to regain directional control if necessary. However, when operating a motorcycle, the operator cannot easily apply both brakes to their fullest. If the front wheel “locks” the motorcycle is going down. If the rear wheel locks, most experienced riders can keep the motorcycle upright, but they are committed to a fairly straight path of travel while the tire is locked. If the front wheel locks, unless released immediately, even the most seasoned and trained rider is going down within a few feet. Therefore, it is very difficult for most motorcycle riders to obtain the optimum braking performance of their vehicle. ABS braking on motorcycles appears to be the system most similar to a passenger car, in that the operator can apply the brakes to their fullest extent and still maintain control of the motorcycle in most situations.
This study is an attempt to assist an investigator in determining the deceleration rate that a motorcycle experienced pre-impact.
STUDY:
Several previous studies were reviewed and their results were compared to a limited study by the authors. The author’s area of interest was the influence that the operator has on the deceleration performance of the motorcycle during emergency braking maneuvers, and if training/experience affected that performance. Three riders with different experience levels rode the same motorcycle (2004 Harley Davidson Wide Glide with conventional brakes) and attempted to stop the motorcycle as quickly as possible, while maintaining control of the vehicle.
The motorcycle was equipped with a Vericom 3000, which was mounted on the rear fender luggage rack. The braking sequence trigger was activated by the motorcycle’s brake light. The riders were instructed to accelerate from a stop, maintain a certain speed, and then bring the vehicle to a stop as quickly as possible to simulate emergency braking.
The tests were performed on the “run-off” area of a drag strip. All tests were performed on the same date, in the same direction, and under similar conditions. The tests were completed within a four hour time frame, so the surface temperature remained relatively the same. The ambient temperature was approximately 75 degrees Fahrenheit.
Three riders, of varying experience and training levels, rode the test motorcycle. Rider “D” was a licensed motorcycle rider for over 15 years, but had little recent riding experience. Rider “D” had received MSF (Motorcycle Safety Founda-
tion) training when he obtained his motorcycle license.
Rider “G” had over 20 years of riding experience since being licensed, had received MSF training and police motorcycle training, and, at the time of testing, averaged about 5000 miles per year on a motorcycle.
Rider “N” had over 15 years of riding experience, was, at the time of testing, assigned to police motorcycle patrol, and had previous motorcycle road racing experience.
(c)2006CrashDataGroupInc
“Comparing the average decelerations of these three riders, the training and experience of the individual riders did not appear to have a significant influence on their braking performance for this limited number of tests.”
SUMMARY:
Rider “D” achieved an overall average deceleration of approximately .65 g on his four tests. His peak deceleration exceeded 1 g on one of his four tests. This rider achieved decelerations from .53 to .73 g’s.
Rider “G” achieved an overall average deceleration of approximately .66 g. Rider “G’s” peak deceleration exceeded 1 g on eight of his ten tests. Two of the tests recorded maximum deceleration of over 1.5 g’s. However, on those two tests where the maximum deceleration exceeded 1.5 g’s, the average deceleration for those test runs were the lowest recorded for that rider. This rider achieved decelerations from .41 to .81.
Rider “N” achieved an overall average deceleration of approximately .62 g. Rider “N’s” peak deceleration exceeded 1 g on five of his nine tests. This rider achieved decelerations from .43 to .74 g’s.
Combining all twenty-three tests, the average deceleration was found to be approximately .63 g’s. The three rider’s deceleration averages ranged from .62 to .66 g’s. While the average deceleration range was relatively small, the individual test run decelerations ranged from .41 to .81 g’s.
In this limited study there appeared to be no significant difference in braking performance between the three riders based on their training/experience. Comparing the average decelerations of these three riders, the training and experience of the individual riders did not appear to have a significant influence on their braking performance for this limited number of tests. Some correlation might be found in a larger number of tests, as in previous studies.
Deceleration averages ranged only from .62 to .66 g’s. While this range is small, the individual test runs ranged from .41 to .81 g’s. It appears that a motorcycle’s braking performance is not primarily based on the braking capability of the motorcycle, but on the braking capability of the rider. It also became apparent that the average deceleration might not be an appropriate value to use when determining the deceleration value to apply to a singular event. Since the deceleration capability of the motorcycle/rider unit is dependent upon the rider’s capability, the rider’s performance in any singular event is the controlling factor for that event. In this study there was a wide range of deceleration values, therefore, the author recommends that a range of values be applied when determining the deceleration in a singular event.
In determining a range of deceleration rates to use in a particular collision, the investigator must consider the rider/motorcycle combination braking performance. Close attention must be paid to roadway evidence, such as deceleration marks. An examination of the motorcycle tires is crucial in determining if that particular tire was braking, or not. A motorcycle tire braking at its fullest capability will sometimes exhibit tire “speckling” from pavement grindings, such as that found on passenger car tires with ABS. The examination of the motorcycle tires, along with the roadway evidence, can help the investigator to determine the appropriate range of deceleration rates that would apply in the collision at hand. The damage sustained by the motorcycle at impact can also assist the investigator in determining the pre-impact braking that was occurring.
1. Braking Performance of Experienced and Novice Motorcycle Riders – Results of a Field Study by K. Vavryn and M. Winkelbauer, Austrian Road Safety Board (KfV), Austria
2. Motorcycle Accident Cause Factors and Identification of Countermeasures, Vol. 1: Technical Report, Hurt, H.H., Ouellet, J.V. and Thom, D.R., Traffic Safety Center, University of Southern California, Los Angeles
3. Braking Deceleration of Motorcycle Riders - H. Ecker and J. Wassermann – Vienna University of Technology
4. Summary of Motorcycle Friction Tests - Bruce F. McNally, Northeast Collision Analysis, Inc.
5. Evaluating Motorcycle Skidmarks - Wade Bartlett, Mechanical Forensics Engineering Services, LLC
6. Motorcycle Sliding Coefficient of Friction Tests - Bruce F. McNally, Northeast Collision analysis, Inc., Wade Bartlett, PE, Mechanical forensic Engineering Services, LLP
Celebrating 70 Years
If you need to learn the basics, brush up on your technical skills, or advance your knowledge, you can count on the experts at Northwestern University Center for Public Safety.
The Center (formerly the Traffic Institute) continues to lead the world in traffic crash investigation and reconstruction. Serious investigators like you come from all over the United States, Spain, China, Africa, the Middle East, and many other countries because they understand how important it is to get it right.
You can trust Northwestern University to deliver the same high quality instruction from the best engineers and reconstructionists, whether you study at Northwestern University’s Evanston campus, at your agency, or online.
-Accident Investigation 1
-Accident Investigation 2
-Basic Physics and Mathematics Workshop
-Vehicle Dynamics
-Traffic Accident Reconstruction 1
-Traffic Accident Reconstruction 2
-Traffic Accident Reconstruction Refresher
-Pedestrian Vehicle Traffic Collisions
-Heavy Vehicle Crash Reconstruction
-Photogrammetry for the Accident Reconstructionist
-Accident Investigation 1 Online
Mostofyouhavebeenusingsimulation(e.g., EDSMAC4, SIMON, EDVTS…)forseveralyears,yettomanythe innerworkingsremainsomewhatmysterious.This TechnicalSessiondescribestheinnerworkingsofa simulation.Wecallthisthe AnatomyofaSimulation
Butfirst,let’stalkabouttheinputsandoutputsfora typicalvehiclesimulation.
Thetypicalsimulationinputparametersare:
VehicleData–Dimensionsandinertias,tire-road friction,stiffnesscoefficients(forcollisionsimulations) andpossiblyotherparameters,dependingonthe complexityofthemodel(e.g.,suspension,steeringand brakingsystemanddrivetraindatamaybeusedby3-D simulations).
EnvironmentData–Aphysicalmodeloftheroad surface.Thismaybeassimpleasaflat,horizontal planeoracomplex,3-dimensionalterrainwithvarying friction.
EventData–Initialpositionandvelocity(minimum requirementforallvehiclesimulations),driversteering, brakingandthrottletables,andpossiblyother parameters,againdependingonthecomplexityofthe simulationmodel.
SimulationControls–Timeintervalsforcalculations andforoutputtingsimulationresults.
Thetypicalsimulationoutputresultsare:
Kinematics(pathposition,velocity, acceleration)
Kinetics(summationofforces,moments actingonvehicle)
Tire(forcesateachtirecontactpatch)
Simulationsoperateinthetimedomain.Thatsimply meansthattheaboveoutputresultsarecalculatedand outputatregular,user-definedoutputintervalsasthe vehiclemovesalongitspathfromitsinitialpositiontoits final/restposition.Thisfactmakessimulationsidealfor visualizingavehicle’smotionduringanaccident sequence.Simulationprovidesthebenefitofbeingable topickanypointalongthevehicle’spathandstatethe forcesthatcausedittobethere(asopposedto animation,whereinavehicle’spositionisatthesole discretionoftheanimator).
Allsimulationmodels,whetheryou’resimulatinga vehicle,ahumanoratropicalstorm,includefourbasic components.Thesecomponentsare:
ControlRoutine
NumericalIntegrationRoutine
PhysicalModel
DerivativesRoutine
ThesefourbasiccomponentsareshowninFigure1.
(c)2006CrashDataGroupInc
Asitsnameimplies,thecontrolroutinecontrolstheflow ofexecution,makinglogicaldecisions,suchaswhenit istimetooutputthecurrentresultsandwhento terminatetherun.Butthemostimportanttaskofthe controlroutineissimplytopassprogramcontroltothe numericalintegrationroutine.
Wementionedthatsimulationsoperateinthetime domain.Giventhecurrentsimulationtime, t, numerical integrationcalculatesthevehiclepositionandvelocity afterasmalltimeinterval, �t.Insimpleterms,the positionattime t+ �t is:
SttSvt ()+=+∆∆
andthevelocityattimet+ �tis
VttVat ()+=+∆∆
where
S =position(typically X,Y and � fora2-Dsimulation)
V =velocity(typicallyu(forward),v(lateral)and � Ψ (angular)fora2-Dsimulation)
a =acceleration(typically u (forward), v (lateral)and �� Ψ (angular)fora2-Dsimulation)
t =currentsimulationtime
�t =calculationtimeinterval(typically0.02seconds fora2-Dsimulation)
Actually,anyonewhohasstudiednumericalintegration knowsit’snotquitethatsimple.Butitisusefultoknow that,regardlessofhowsophisticatedorcomplexthe numericalintegrationroutinemaybe,theprocessboils downtotheprocessdescribedbytheaboveequations.
Weprovidedtheinitialposition, S,andtheinitial velocity, V, aspartofourinputs(seeEventData, above).Weprovidedthetimeinterval, �t,aswell(see SimulationControls,above).Butwheredoesthe acceleration, a,comefrom?Forthat,wenextrequirea vehiclemodel.
Thevehiclemodel(seeFigure2)isdefinedbythe vehicle’swheelbaseandtrackwidth(actually,the wheellocationsrelativetothevehicleCG)andits weight,W,andyawrotationalinertia, I Ifthe simulationincludesacollision,exteriordimensionsand stiffnessarealsorequired.
Thevehiclemodelalsoincludesatiremodelthat calculatestheforcesactingateachtire.Althoughtire modelscanbequitecomplex,thoseusedby2-D simulationsarerelativelysimpleandrequiretiredata (corneringandfrictionproperties),tireslipangleand verticaltireload.Slipangleandverticalloadmustbe calculatedforthecurrenttimestep.Theoutputsfrom thetiremodelarethe F and F forcescurrently producedateachtire.
- -
Ifacollisionissimulated,acollisionmodelisalso required.Althoughthereareanumberofwaysthiscan bedone,acollisionmodelalwaysboilsdownto calculatingtheforceactingonthevehicle’scrushed exterior. F and F crushforcesarecalculatedat specifiedpointsalongthedamageprofilebasedon Hooke’slaw:
Fkx = where
k =vehiclecrushstiffnesscoefficient
x =displacement(crush)
Otherforcesmayalsobemodeled(e.g.,aerodynamic forces,connectionforcesfortrailers).
Regardlessofthenumberandtypeofforces considered,theoutputfromthevehiclemodelisthe totalsumofallforces(tire+collision+…)actingatthe vehicle’sCG,andthetotalmoment(rotationaltorque producedbyalloftheseforces)abouttheCG. Mathematically,wehave:
ΣF =sumofforcesinxdirection
ΣF =sumofforcesinydirection
ΣM =sumofmomentsaboutz(vertical)axis
Thefinalstepiseasy.Giventhetotalforcesand momentactingonthevehicle,alongwiththevehicle mass(weight/g)androtationalinertia,weuse Newton’s2 lawtocalculatetheaccelerations.In general, � F=ma .Wecanrearrangethistogetthe forward,lateralandangularaccelerations:
Ψ =angularaccelerationaboutthevertical(z)axis
Oneminorcomplication:Becausewecalculatedthe accelerationsinarotatingreferenceframe(i.e.,the vehicle),weneedtotransformthelinearaccelerations backtoaninertialreferenceframe(i.e.,theearth):
(Angularacceleration, Ψ,isthesameinbothreference frames.)
Theaboveaccelerationsareusedbythenumerical integrationroutine(referbacktothelastpartof NumericalIntegration,above)tocalculatetheposition andvelocityattheendofthetimeinterval, �t.Sincethe vehiclecanmoveinthreedirections(X,Y and �),we needtocalculatethreenewpositioncomponents:
Wenowhavethepositionandvelocityofthevehicle aftertimeinterval, �t (seeFigure3).Sowearedonefor thecurrenttimeintervalandwe’reready(i.e.,wehave theinitialpositionandvelocity)forthestartofthenext timeinterval.Controlreturnstothecontrolroutine(see Figure1).
ReferringagaintoFigure1,wehavereturnedtothe controlroutine.Thecontrolroutinechecksfor terminationconditions,timeforoutput,aswellasany otherissuesrelatedtooverallprogramflow.Unlesswe haveaterminationcondition(whichcausesthe programtoexit),thecontrolroutineagainpassesthe batontothenumericalintegrationroutineandthe processcontinues.
Akeyobservationisthatthecalculationsequenceina simulationprogramisveryrepetitive.Theabove processcontinuesinaloop,marchingforwardduring eachcalculationtimeintervaltoupdatethepositionand velocityofthevehicleuntilaterminationconditionis reached(usuallythevehiclecomestorest).
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
In the first issue of Collision, we offered a case study based on a crash test conducted using a Cavalier and a Lumina. As part of that problem, we posed this query: “The Lumina (CDR system) download shows a non-deployment event. That non-deployment event shows a delta-V of -1.1mph over about 34ms. For the sake of this case study, one may assume that the actual, total delta-V for the Lumina is about -10.3mph. As it relates to this case study, assume that this value may have been calculated as part of a reconstruction. Explain the difference in the calculated delta-V of -10.3mph and -1.1mph as reported in the Lumina data report.”
What we’re saying was: we know from an analysis (and in this case, from instrumentation because it was a documented test) that the real, total delta-V for the Lumina was about -10mph. The EDR component in the SDM captured a delta-V of about -1mph...why such a marked difference?
The solution offered was: “Now, it’s time for a more thorough analysis of the crash. We need to do a more thorough momentum analysis for impact and closing speeds, total delta-V and then PDOF...but those are the clues. The more complete answer in the next issue of Collision where the topic of this section turns to calculating closing speed!” (Alright, it was a non-solution and yes, it was a “cliffhanger” - no different than a newspaper that gives you the answers to the crossword in the next edition...)
Well, now we are into issue 2 of Collision (you’re reading it!) and in this issue we’re going to address closing or impact speed calculations in the current Case Problem but, more directly to the point of this question, this issue of Collision includes a great explanation of PDOF ... and THAT’s the key - or at least part of the key - to addressing the clearly underestimated delta-V for the Lumina.
THE FACTS:
In the last issue, we detailed the crash test. The last Collision included photos, a scene diagram, a description of the test and answers to all but this last question. Since this was a test, we know the impact speed/ closing speed of the Cavalier into the stopped Lumina, we know the actual delta-V of the Lumina to be about -10mph but that value captured at the Lumina SDM shows up as about -1mph. No getting around it, that’s a remarkable difference. The question remains: why? Is it because data from an SDM is just no good? Is it because it’s “just delta-V” and we all “know you can’t do anything with delta-V?” Well, the answers are: (1) mostly because of PDOF, (2) no, and (3) What, are you out of your mind?
THE ANSWERS:
There are actually two parts to the answer to this, one is fairly easily quantifiable and the other requires an understanding of the functional capabilities of the Lumina SDM. Inasmuch as we just had a lesson in PDOF, let’s take that approach first. In the underlying crash test, the Cavalier approached the Lumina at an angle to the stopped position of the Lumina (using, for the moment, the Lumina “X” axis (see SAE J211) as a frame of reference). Moreover, and this observation leads to the second part of the analysis: there are limits to the SDM’s ability to capture acceleration as a function of its design and purpose.
(c)2006CrashDataGroupInc
In the underlying crash, the Cavalier hit the Lumina on the left (driver’s side) almost fully “broadside” into the driver’s side, more toward an approach angle of 90degrees (see SAE J211) than 0 degrees - relative to the Lumina - and well behind the Lumina’s center of mass.
As we review the Russell analysis of PDOF, the answer becomes more clear, the effect of this collision isn’t falling on
- Brad Muir, Rusty Haight and Dan Miles
the Lumina’s X axis and, since that car has an accelerometer in the SDM that only “looks” for negative X axis acceleration - and from that we get the reported delta-V - the Lumina wasn’t “seeing” much of the effect of the crash, but that’s the second part of this analysis. In short, the PDOF for the Lumina, for this crash, is far greater than 0deg (relative to the car’s baseline X axis) and we can, to a certain extent, quantify the effect of that off axis force application. (see PDOF article at page 15 in this issue of Collision)
Now, a review of the available information tells us that the approach of the Cavalier to the stopped Lumina creates a PDOF, for the Lumina, of about 280-285 degrees (see SAE J211). Using the Lumina’s X axis as a frame of reference, that means that the effect we should see on the X axis would be about -1.8 to -2.6mph (here, we’re using cos(theta) (-10.3mph)).
Contemplating the potential for the PDOF to have been between 275 and 290deg (a wide range, but for this example it may be helpful), we can calculate a range of values showing what portion of the total -10.3mph delta-V would have been observed on the Lumina’s X axis. If the PDOF were to be the angle theta and we evaluate the range of angles, we find the resultant which would be observed on the X axis could fall between -0.85mph and -3.5mph. Calculating the values in 5degree increments, we might see the calculated X axis effect develop as:
PDOF
Calculated value (based on a total delta-V of -10.3mph)
275deg -0.89mph
280 -1.78mph
285 -2.6mph
290 -3.5mph
Carrying the values to 270degrees, a fully broadside crash, would, of course, leave us with a value of 0mph on the X axis.
Immediately, one might see that there appears to be a linear trend developing where there’s an average of about0.88mph between the incremental steps. Assuming the trend to be linear, that would seem to run counter to the
information found in the Russell PDOF article in this issue. Russell showed us that Fnet is proportional to Fx in a nonlinear way. This is because the proportionality multiplier “k” in Fx = kFnet is not linear or constant when k = cos(theta) as it’s depicted here.
However, on a narrow interval of 270 to 290 degrees, the cos(theta) function behaves in a fairly linear way. If we enlarge the line segment in Figure 1 from “a” to “b” we can see that at least a portion of the cosine curve between those angles is fairly close to line segment “a”-“b” which would have a slope of about -sin(280). Since the derivative of cos(theta) is -sin(theta) we can use 280 in our -sin(theta) function to approximate the derivative because it is approximately the mean angle theta in our angle interval.
The reason that Fx ultimately changes by a different value in between the given points is that the cos(theta) function is decreasing at an increasing rate. This results in a greater and greater decrease in Fx at every angle increment. The fact that our angle is only changing by 5 degrees over the range means that cos(theta) will only change by a small, almost negligible amount which explains the very small differences in the change in Fx.
In a nutshell; however, evaluating the PDOF even across a fairly large range wouldn’t explain, completely, the differ-
ence between the calculated delta-V and the delta-V reported by the SDM and even looking at it that way, we’re still under reporting the delta-V at the SDM, right? That leads us to analyze the function of the SDM. The SDM - Sensing and Diagnostic Module - is an airbag control module. It is not a “black box,” an “event data recorder” or an “evil tattletale installed by some vague “big brother-like” spy agency.” The SDM is installed as a unit which runs a diagnostic routine on the frontal supplemental inflatable restraint (SIR) system (the frontal collision airbags and related components for this car) and then, when conditions are right, it captures some of the information it had available to it (specific to this example, acceleration) at around the time of the crash and stores it internally for later retrieval and review and that is the delta-V we find in the Vetronix CDR report.
Further, that module has to experience what’s known as “algorithm enable” or “wake up” to even begin analyzing acceleration data to ultimately make that deploy - don’t deploy decision.
So, the second half of the answer to last issue’s case problem and that which would further explain the difference we still see after the adjustment for the PDOF - although much less in magnitude than we see in the raw numbers themselves - is that (a)the module had to “see” some negative X acceleration de-
velop just to wake up and start evaluating the crash and (b) we know from other research and published works that at lower delta-Vs (as experienced on the X axis) there is the propensity (as a function of wake up) to underreport delta-V.
So, at the end of the day, there really isn’t so much an “error” or “flaw” in the delta-v reported by the SDM as there is a need to evaluate the data reported in light of the facts of the case at hand and, doing that here, we see what caused this “difference” and can explain it.
Cos Graph
For years accident reconstructionists have been using aerial photographs in their technical crash analysis. Many existing GIS mapping systems used high-resolution B&W orthogonal photographs as the basis for their early CAD drawing of Governmental Assets. These early conversions of raster images to CAD images for county and city engineering operations were one of the first moves from paper based hand drawn plans to computer-based drawings.
In order to use these images the early users had to be sure that each image had been properly rectified to be sure that it was an accurate and true orthogonal image. These images were then imported and the desired line-work was traced in the CAD environment. After this was done there were various scaling techniques that allowed the user to be sure that the objects in the drawing were accurate and scaled to real world size.
Today there are many warehouses of images currently available to anyone who would like to use orthogonal images to create their own maps of roadways, buildings, sidewalks and other items that are useful to faithfully create a plan view map of a crash scene. Most of these images are available on a pay for use basis and the resolution and quality of rectification varies widely.
In the late 1990’s a company named Pictometry began to develop a system by which the company would image an entire city, county or state after it contracted with a municipality. The company develops and markets a sophisticated, integrated information system that allows users to have high-resolution images of neighborhoods, landmarks, roads, and complete municipalities from multiple viewpoints. A specially equipped plane with digital cameras mounted in both oblique and or-
thogonal orientations would essentially “mow the lawn” of a given area while computer and GPS equipment controlled the operation of the camera equipment and gave guidance to the pilot for his flight path. Along with each image taken there was a packet of data that included time, date, GPS positioning data and information on the pitch, roll and yaw attitude of the aircraft at the moment the image was taken. This results in fully georeferenced images that can be used for many different applications.
These images should not be considered “survey grade” as onscene physical measurements and measurements with total stations and similar laser based instruments are always the most accurate and reliable method. But when additional data and items are needed use of these images is many times the only way to go.
Pictometry images include all the data necessary to use them without any prerequisite knowledge of coordinate, datums or projection systems. These digital aerial images allow the user to see detailed information and land attributes such as buildings, roads, trees, sidewalks, light poles and pavement markings.
The images were then processed and a virtual mosaic of digital images of an entire municipality would be created using the proprietary Pictometry software. As a result of their work, Pictometry is the world’s largest digital, aerial photography company.
For accident reconstruction the most usable type of image from Pictometry is the Neighborhood Orthogonal. These images cover roughly 1100’ by 1600’ in size and have a resolution that produces 6” pixels. The image has its best accuracy in the middle of the image and the accuracy naturally trails off as
you move toward the edges. The best way to use the large-scale high-resolution orthos is to either export it from the Pictometry software, or purchase the rights to use the images you need directly from the company.
When an image is exported from the software it is best to save it as a .jpg as opposed to a TIFF (figure 1). When the image is saved to a location on your hard drive a companion data file is created (figure 2). It is best to copy the data available in the software and paste it into this file to make scaling easier later (figure 3).
The resolution of the stock Pictometry images can be can increased by using a process called Posterizing that is described by Scott Kelby in his book The Photoshop Book for Digital Photographers. By increasing the size of an image by 110% there is an increase in size that is proportional to the percentage of increase, but the relative sharpness of the image doesn’t decrease relative to the new size. This percentage increase can be done multiple times with an ever increasing size and decreasing relative sharpness. The author attributes this routine to Jim DiVitale, and by doing this relative resizing multiple times you can greatly improve the quality of the image before using it for the creation of your map or map elements. However, the user needs to be careful as some CAD programs limit the size of images to be inserted to 10 MB or so. The user can also adjust the color and contrast of the of the image and even convert it to a grey scale image if that makes it easier to work with before saving the changes in Photoshop and preparing to use it in a CAD environment.
There are several ways to insert a raster image into different CAD programs and each has their own way of dealing with original file and how it is displayed. The older version of MapScenes Pro requires that the image be inserted as a background. Newer versions of AutoSketch (v-5+) allow the user to insert an image on it’s own layer, but images inserted directly take a long time to load. If the image is simply linked on the insert object option box the image won’t be found if the file or image is moved to a storage card or other computer.
AutoCad, MapScenes Pro 06 and other cad programs that are true 3-D programs allow the user to insert the image on it’s own layer as an object. This object insertion routine makes it easy to turn the image on and off and move it into the background by ordering the layers. Scaling of an image inserted as an object is straightforward by comparing the properties of the inserted image to the data saved with the exported image from the Pictometry environment. The user only needs to divide what the real world size by the dimension of the image after it has been inserted. By using the windows calculator the scale factor calculated can be copied and pasted directly into the input box in AutoSketch or AutoCad. MapScenes 06 makes scaling simple because the Insert Raster Image dialog box default scale factor is 1 for both the X and Y axis. You simply enter the longer “real world” dimension from the Pictometry data file and you now have the image in real world size.
Once the image is inserted and scaled save the file immediately to be sure the path to the file is preserved. To avoid linking problems and avoid losing the image it is best to create a file folder and use it to store the CAD file and original image. This will allow the user to simply copy the entire folder if it needs to be moved. Once this is done you can begin identifying and tracing the elements you need to create or complete your map.
Figure 4 shows the image with the linework for the new store and parking lot and the pavement markings. Figure 5 shows the linework with the image turned off. Figure 5 shows the finished plan view map of the scene after the image-generated linework had been added to the GIS and the total station data from the scene.
The user must remember that organization is everything when drawing on a computer. For that reason it is a good idea that you create a blank or template with all the necessary layers already created. Before saving this blank make the image layer current then when you save and re-open it will be ready to go. By knowing what you have before starting to measure any scene you can save time by only measuring what you need that is particular to your crash. Always remember to include at least three redundant points to assist in inserting your scene measurements to either your GIS or any map created by using Pictometry images. Light poles make excellent redundant points as they are generally straight and are easily identified at the base by the shadows in orthogonal images. If you are using building corners as redundant points pay close attention to the overhang of eves and other roof elements. The image and GIS maps will include the overhang, but measuring at the base of the building foundations will show a smaller building footprint.
Another useful application of this method is to examine the scene photos post investigation and identify points of interest and compare their location to man hole covers, pavement markings, crack seal and naturally occurring defects in the roadway surface. Then you can identify, locate and draw these elements on the ortho image and actually get reliable locations of crash scene elements not measured at the time of the on-scene investigation.
SUMMARY:
By using large-scale aerial orthogonal images crash reconstructionists can greatly improve the detail included in their scene diagrams. By inserting these images into their CAD program of choice many details that are difficult to document on an open roadway can be included in their diagrams.
Although this method will never replace the need for quality and timely measurements taken directly from the crash scene, you can add many elements using this imagery, and make your scene maps much more detailed.
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
CLawyers Association, considers it “devastating evidence,” saying, “if you have a good animation, it’s such a difficult thing for the other side to fight.” A visual account of the incident need not be played to the jury to have an effect in litigation; many times it precipitates a settlement before a trial begins. In the six cases in which Golomb has used computer animation, opposing counsel settled almost immediately.
Legal animations, a combination of artistic effects and careful scientific investigation, originate from accident reconstruction firms like the Denver-based Knott Laboratory, which has performed over 10,000 accident analyses over its 24-year history. For the last decade, its animation department has placed the mathematic conclusions of engineers into the visual language of CGI.
The company’s website, www.knottlab.com, shows short clips of past cases. In this mini-cinema of disaster, the main characters are machines: a BMW careens into a tunnel wall; a construction crane teeters before toppling, and a top-down convertible speeds obliviously into the path of a locomotive. The fatal errors of these mechanical characters are often a matter of inches, and the destructive consequences occur within the span of a few seconds.
Debate over exactly what happened during this small window of time may sprawl into many years of litigation. The Knott Lab animations give attorneys the power to define the basic facts without ambiguity. What may be only a flash in the memory of an eyewitness, can be shown in slow motion and from any angle. Knott Lab can accurately portray the incident even if no one, in reality, saw it happen.
“If they placed video cameras at every intersection around the globe, and videotaped every accident, they wouldn’t need us,” says David Danaher, P.E., Director of Mechanical Engineering at Knott Lab. “Unfortunately, we don’t enjoy that luxury.” While their realistic effects are recognizable, created from the same 3D technology used in the movies for entertainment, what makes these short clips non-fiction? Besides animation tools, Knott Lab utilizes another 3D application -- a software called PhotoModeler.
KThrough just a few photographs of the aftermath and the small traces of evidence, investigators are able to reconstruct and render past events. PhotoModeler from Eos Systems requires only three photographs of an object from different vantage points for enough geometric data to construct a 3D model, including very accurate dimensions of everything within a photographic scene.
To virtually choreograph an auto accident, Danaher and his engineering team work backward, analyzing the final rest positions of vehicles, debris, tire marks, and most importantly, the extent of vehicle damage. The metal folds in the auto body, referred to as the crush, hold the key to many of the dynamics of an accident, including the speeds, locations, and angles of impact.
Much of the primary evidence, including the crush, can only be found in the accident-scene photographs taken by police or insurance agents. If a crushed vehicle can be located and recovered, the Knott Lab investigators take measurements directly. Scrap metal, however, often moves faster than the wheels of litigation. Fifty percent of the time, Danaher says, the vehicle is “long gone” by the time Knott Lab takes on the investigation. The junked cars pass between insurance companies and salvage yards.
“We need to know how much damage is done and we have to have a valid scientific proven method to quantify the damage to the vehicle without physically seeing it,” explains Danaher. “We’re really limited in what we have for evidence. If we have three or more photographs of a car, it’s cheaper to use PhotoModeler rather than sending a team to find the car and survey it.”
PhotoModeler automates the science of photogrammetry, which means literally, “measurement from photographs,” a technique that has been around as long as cameras. Over the last decade, the science has gained acceptance legally as an evidentiary tool. “In our cases, we undergo Daubert challenges, which are very strict federal gate-keeping procedures for expert testimony,” reports Director of Animation Hailey Day, “PhotoModeler has consistently held up.”
On a stretch of I-95 in Florida, a semi tractor trailer struck a Ford Expedition, causing the SUV to rollover several times, killing the driver.
When litigation ensued years afterward, Knott Laboratory was tasked to determine and animate the sequence of events. The first issue concerned the roof of the Expedition, and whether any liability lay with the auto manufacturer.
“Here the defense is essentially arguing, ‘although [the semi] may or may not have caused the accident, it didn’t cause the death, because her vehicle was constructed poorly. If it hadn’t crushed the way it did, she would have survived,’” explains Danaher. “The question was: should the vehicle have been strong enough to sustain this rollover?”
For a scientifically valid answer, engineers calculated the force that caused the roof damage by careful measurement of the deformation. PhotoModeler accurately dimensioned the damaged Expedition, generating a three-dimensional outline of the vehicle, called a spline cage.
Because the team was able to recover the SUV in a salvage yard, investigators could also verify these photogrammetric measurements through a laser survey. “Other than a few small discrepancies,” says Director of Animation Hailey Day, the spline cage from PhotoModeler and from the laser measurements were “a dead-on match.” (Figure 1).
In producing 3D models for analysis and animation to (Figure 2), the photogrammetric method is in some ways preferable, since it captures details all at once, which otherwise would require many small linear measurements. Senior Forensic Animator Thomas Reyes: “In any case we do, we try to take complicated damages that are very three-dimensional, that you can’t just measure with
necessary to make the indentation.
“The maximum crush -- right around the driver’s head --was about 19 inches. It was a huge displacement,” says Day (Figure 3). Once the photogrammetry is complete, Knott Lab engineers compare the displacement data to that of federal safety standards.
Investigators also need to explore another issue in the incident. Just prior to the collision with the Ford Expedition, there was a minor sideswipe between the semi and a small passenger car, a green sedan. If the engineers can prove the green sedan was at fault in the initial contact, then the trucking company would not carry the full weight of negligence.
as a reliable
a tape measure. We use PhotoModeler to get a better 3D model.”
Reyes imported the spline cage into a rendering and animation application (3D Studio Max by AutoDesk) and added the photorealistic details to the model. Danaher and the engineering team compared the crushed model to a computer model of an undamaged Expedition. Given the comparison, along with industry data about the strength of materials, the engineers could find the force
Police had photographed the sedan’s side-door damage and interviewed the driver after the incident. Photogrammetric analysis on the sedan indicated that contact had obviously occurred, but the
facts were unclear as to whether the car or the truck was at fault for this first collision. “Everyone agreed on the basic premises,” says Danaher. “The real question was: ‘who hit whom in which lane?’”
Usually in accidents involving a major impact, the road reveals clues to the position of vehicles. Asphalt gouges, tire skid marks, and fallen debris give engineers strong points of reference to determine the location of the collision. Police departments who use photogrammetry in accident cases capture these small details as a matter of procedure, so a highway scene can later be mapped for trial.
“What we didn’t have is a contact point between the semi and the sedan, because there was very slight contact between the two of them, and obviously that doesn’t leave a big gouge in the roadway,” says Danaher.
Instead, the clues to the lane positions came from inside the truck itself. “The semi had what you might call a ‘black box,’ which records data,” recounts Danaher. Unlike the aviation flight recorder, the truck’s onboard computers were not intended to explain events of a crash. A GPS unit continually reads the geographic coordinates of the rig, so the company can track the progress of loads. Computers also monitor the mechanical health of the truck by tracking the speeds, turns, and brakes. For safety, a radar system senses and tabulates the presence of other vehicles on all sides. With this trove of digital information, Knott Lab was able to correlate the data in context of the road during the time of the accident (Figure 4).
With firm evidence of the truck’s path, the team was able to form a confident opinion that the sedan merged into the truck’s lane. Animations produced from 3D Studio Max could then show the full chain of events (Figure 5).
“The case ended up settling before going to court. It never went to trial,” says Danaher, “Our client was very happy with what we did.”
Developments in 3D technology offer attorneys not only a powerful visual aid, but also the means to determine the crucial facts of the case. With results that are scientifically valid and legally admissible, photogrammetry tools and animation allow Knott Lab to reverse-engineer reality.
Founded in 1982, Knott Laboratory provides forensic engineering and computer animation services to comprehensively reconstruct accidents. Knott Lab has worked on more than 10,000 failure-analysis cases nationwide for the legal system and the insurance industry, as well as for local and national businesses. Knott Laboratory assist its clients from initial contact through investigation, analysis, animation production, and expert witness testimony. The firm’s expertise includes vehicle accident reconstruction, product liability, mechanical engineering, civil and structural engineering, electrical engineering, fire and explosion investigations, computer animations and graphics.
Eos Systems Inc is the developer of the award-winning PhotoModeler software and is the leader in versatile close-range photogrammetry solutions. PhotoModeler provides an easy and affordable solution for measurement or reverse engineering of objects into 3D CAD through the use of photographs. The software is used by thousands of companies specializing in crime and accident reconstruction, archeology, architecture, engineering, surveying, film and video animation, among others. Eos Systems is headquartered in Vancouver, British Columbia.
It is often thought that the sure route to fame and fortune is the publication of a successful book or series of articles. You’ve heard the spiel no doubt – “Write that book, get invited to talk about it with a host or hostess of a nationally syndicated talk show, go on a book-signing tour, and cash in that terrific six-figure publisher’s advance…” If only it were so easy!
The facts are anything but easy. The majority of the income derived from authoring books and other publications is earned by a very small percentage of all the authors who manage to get into print. The vast majority of published writers collect enough rejection slips to wallpaper their living rooms and bedrooms combined. Large royalty advances rarely go to anyone other than established authors with proven track records of successful book sales or to famous individuals who can cash in on their momentary fame or notoriety. Writing is hard work – really hard work for the novice writer. If this were not so, there would not be a market for the hundreds of “ghost writers” who make their living assisting inexperienced authors through the process of getting published.
Nevertheless, there are some tangible benefits to being a published author. This is especially the case for experts courting advanced standing within their chosen specialty or profession. Such benefits include wider name recognition, speaking and teaching opportunities arising from the increased exposure gained by being a published author, and the enhanced status that comes with being perceived as a highly regarded expert competent enough to have published on the subject in question. In addition, there can be a true sense of self-affirmation associated with the act of creating a book or article on a subject near and dear to the author’s heart.
Many published authors speak fondly of the positive associations they enjoy with other successful writers and with their readers’ expressed appreciation of the author’s published work. Ultimately, the published author takes advantage of the opportunity to create a legacy of sorts by contributing his or her written record to civilization’s cumulative consciousness regarding some aspect of the author’s passion or life work.
Nevertheless, it is only fair to point out that there are also some valid reasons not to bother attempting any published writing. At
the top of the list is the fact that for most writers, actual compensation achieved is lower than minimum wage for the amount of the author’s time and work expended in researching, writing, and promoting the published work. Add to that the fact that for an expert who makes frequent court appearances as a professional witness, published documents can spell trouble on the witness stand should a written record turn out to be at odds with a current position being advocated by the expert. On top of this, the increased exposure that comes with being a published author makes the expert a better “target” on the witness stand. The author’s written words provide genuine clues concerning attitudes, research styles, investigative habits, and admired conduct. Should an expert dare to approach a real life problem in any fashion less exhaustive than what the author declared in his writing to be the “proper” way to problem solve, it is inevitable that sooner or later a sharp attorney will catch the difference and broadcast his observation to the judge and jury in less than flattering language. Ouch!
Any way one approaches this issue, at the heart of the debate is the effort that must go into becoming a published author. Writing –effective writing – is difficult and time consuming. The polished writer spends a great deal of time re-writing and re-wording, consumed by his or her desire to produce the perfect product. The inexperienced writer spends no less amount of time, if for no other reason than the fact that the writer’s lack of experience creates barrier after barrier related to “learning the hard way” as he or she trudges through the minefields of the novice author’s struggles.
Since the title of this article suggests the dispensing of advice related to getting published, it is proper from this point on to assume that anyone reading this article has considered the good and bad of authorship. If you keep reading past this sentence, it is assumed that you have made the decision to move ahead in some fashion towards a goal of becoming a published accident reconstruction author. Congratulations! Grab an extra cup of highly-caffeinated coffee and get out your colored highlighter.
What follows is a quick tour of the essentials in getting published, revealed by one who is both a published author and a publisher of other authors, an avid reader and a tormented writer. (The tormented part comes from having a father who was an English
teacher. Said Father insisted on my learning proper sentence construction prior to mastering the more important things in life like fishing, hunting, baseball, and dating.)
What is it that you want to write? The first essential step in becoming an author is evaluating what you have to say in print. Do you want to write about a subject that can be covered adequately in an article? Or does your passion for writing extend to a broader subject that would best be developed in a book format?
Many experienced writers find it helpful to start with a detailed outline. This is constructive for several reasons. First of all, an author who cannot lay out his or her writing idea in an outline format will find it next to impossible to organize the writing well enough to create a piece that will flow coherently for the reader. Second, the outline is an essential discipline for the writer. It helps the writer capture the main ideas of the project, while it also forces the writer to think of the supporting points that develop the thrust of the main ideas.
Outlines are also useful as an evaluation tool in terms of estimating project length. Many a prospective author has thought for years about writing a book, only to sadly discover midstream that the sum total of the written work was really only adequate for a short story or an article.
The use of an outline becomes even more important to the development of the writing project when the writer stops to consider how best to create a working sequence of ideas. Getting the ideas down on paper and rearranging them as needed for proper flow not only helps with organization, but it also assists the writer in avoiding the dreaded “writer’s block” that manufactures barriers to creativity. Having an outline that serves like a detailed road map is a certain means of driving the “writing bus” from start to finish,
picking up the ideas and points scheduled in advance at each “stop” on the outline.
The second important step is researching the best venue for the writing project. This can be intimidating for novice authors. The following suggestions may be helpful here:
•A great place for “1st-time” authors to start is writing short articles for association newsletters. Virtually all professional organizations affiliated with a given profession publish newsletters for their membership. And almost all such publications scramble for good content on a regular basis.
•Similar opportunities exist for “parallel” organizations, like attorney, paralegal, and insurance claims groups looking for articles. Any professional group that might have some link to the services offered by the expert is a potential writing venue.
•Another relatively easy market to crack is the “in-house” newsletter you can create as a business organization for communication with clients and prospective customers. If your writing actually provides useful training and information of value to the intended readership, this venue may prove to be one of the most cost-effective writing efforts you can attempt. Benefits of this kind of writing include continuing name recognition among prospective and current clients, the repeated reminder of the expert’s subject mastery, and the valuable practice gained by the author in writing and editing professional quality literature. Word processing and publishing software easily available at office supply stores contribute to the ease of creating eye-catching copy that will make the expert appear to be a notch above the competition. Figure 1 is an example of such copy. Figure
Additional venues to be considered include scholarly peer-reviewed papers such as those published by SAE. Not to be forgotten are the trade magazines and journals that are directly associated with the expert’s field, such as Accident Reconstruction Journal or Accident Investigation Quarterly, or the new Collision, The International Compendium for Crash Research. Also important as potential venues are the trade magazines that are indirectly associated with the accident reconstruction expert’s work such as American Trial Lawyer or Claims Magazine.
If one decides to write a book, two popular options are worth considering. The first is courting an existing publisher known to market books dealing with subject material similar to what the aspiring author has in mind for his or her project. Second, the novice author can elect to try the self-publishing option. This will normally mean either paying a “vanity” press to publish and market the book, or forming a publishing company in order to self-publish (including everything from writing, to editing, to printing, and distribution). Assuming that the prospective author has determined the direction he or she would like to go with the writing project, it is time to get serious about crafting a proposal to sell the idea to the chosen publisher. It is imperative that the writer becomes intimately familiar with the subject matter and writing styles of the venue chosen for the proposal. Countless writing proposals are rejected for reasons that have nothing to do with the quality of the writing, but are directly related to inappropriate submissions for the target audience associated with the chosen venue. If the writer becomes comfortable with the publication and its target audience, the writer will have a much greater chance of getting into print.
Next, the writer must make certain that he or she can briefly and concisely describe the proposed writing project and the facts demonstrating why the presentation will be worth publishing.
Step Number 3: Contact the publication for instructions on how to submit a writing idea. Many publications have free materials available for prospective authors that list the specific instructions necessary for successful submission of a writing proposal. Failure to follow these instructions to the letter is certain death for the aspiring author.
Step Number 4: After researching the publisher and its author submission requirements, query with a well-researched and polished proposal letter. The days of non-solicited manuscripts are pretty much a thing of the past in the publishing world. A good proposal must leave no doubt in the reader’s mind about the intended scope of the proposed project. It must also define the intended market for the project, identifying not only who would be likely to read and/or buy the proposed writing, but also how the project will be unique or new in the treatment of the subject matter. Also helpful in the proposal is an indication of why this project will be more successful than competing project proposals from other prospective authors.
Language used in the proposal must be both precise and concise. If the reader does not immediately understand what is being proposed, he will lose interest and enthusiasm in a way that will quickly kill any chance of success. Also, the aspiring author must not only sell the idea that he or she is pitching in the proposal, but the proposal must convince the reader that the author is the best writer
available for the project idea.
An effective proposal will reveal enough about the author to inspire confidence in his ability to successfully complete the project. Publishers “role the dice” with each proposal they agree to consider for publication. As such, they need to find within the proposal reasons for believing in the aspiring author. Thus the good proposal will sell the author and his attitude as much as it sells the project idea. Below are six questions that may be helpful in defining and advancing a proposal idea:
1.How would you describe your proposed manuscript?
2.What will be the over-all theme of your manuscript?
3.Do you have a title in mind for your project? If so, what is it?
4.How would you describe your target audience?
5.Can you provide an outline of your manuscript project, including chapter or topic headings?
6.Can you persuasively describe your qualifications to pub lish on your chosen topic?
Moving Forward with the Project… Let us assume that you are a qualified author and that you have created a fine project proposal that shows you are ready to move forward with your book or article. For the sake of discussion, this article will address the remaining steps in book creation. But everything said will apply to shorter writing projects as well. It is important to recognize that crafting quality pieces that are by nature short in length can be more difficult than writing a book-length manuscript. It takes a superior writer to condense without losing content. Most experts have vast amounts of knowledge, and it can be difficult to take that large amount of information and sift it successfully so that writing can be tight enough and concise in a manner suitable for a short piece of work. In theory, this is like trying to pack fifty pounds of stuff into a ten-pound sack!
To start the process, you will make a major decision about the means by which you believe you can best get into print. Basically, you will either shop your project idea around with an established publisher, or you will gather up the necessary courage and funds to try the self-publishing route. Each method has its advantages and disadvantages.
If you elect to go with an established publisher, you will find that there are a number of advantages commonly thought to reside with this option. The major advantage is that you can concentrate almost exclusively on writing your project. This is a real relief for most authors, since the writing, re-writing, and editing associated with putting together a readable and saleable book are harder than most folks realize.
Another advantage is that the aspiring author bears no financial burden for the book once the publisher accepts his or her manuscript.
Many people feel there is enhanced prestige in being accepted to write for an established trade publisher. Getting a manuscript through the steps of acceptance, writing, and publication implies that:
1.For one, the idea had enough merit to justify acceptance in a competitive marketplace.
2.Second, the project had to be conceptually sound in order to make it through the vigorous editing standards of the publisher.
3.Third, an established publisher just wouldn’t waste time and resources on a “bad” book idea.
But there are disadvantages in publishing with an established publisher as well. Remuneration is typically low. Most trade publishers pay only 8 to 10 percent royalties on gross sales, minus buybacks and remands. In specialty markets that typically involve initial book runs of 500 to 1000 copies, this means relatively low compensation for the author, despite his or her labor of love in creating the book.
Publishers essentially “own” the work once the project is accepted for publication. The contract between author and publisher is heavily weighted in favor of the publisher. As long as the contract is in force, the publisher will typically own the copyright on the work and also controls most (if not all) of the reprint rights associated with the work.
The publisher will also be in control of advertising for the book. This means that the book may not get the exposure the author feels is necessary or appropriate for a successful publication run. And aside from free copies negotiated in the contract, the author must buy the published book from the publisher (though usually at a discount price) if he or she wishes to have additional copies of the book to distribute.
For an expert wishing to get greater exposure in his or her field, money is probably not the primary factor in creating the desire to get published. Because of this, most experts choose to have their work published by established publishing companies. There are some, though, who wish to investigate the possibility of self-publishing. Let’s take a brief look at the options and steps associated with this publishing venue.
One can contract with a “vanity” publisher such as Vantage Press to self-publish. This is much like going with an established trade publisher, except that the author contracts with the “vanity” publisher to accomplish all the steps necessary to get a book into print. The critical difference is that the author pays the publisher for every step of the process, from editing to printing to distribution. Expensive!!!
In today’s digital age, with the wonderful ability to sell and distribute via Internet exposure and sales, it is possible to be a successful self-publisher and even limit start-up costs by way of “Print-onDemand” technology. But there are steps that must be followed in order to succeed.
First, start a publishing company that can establish a brand name. This can be as simple as DBA under an assumed name certificate. Once the publisher has an established legal identity, it is then possible to move forward with the registration of a work in print. This means the acquisition of ISBN numbers for each work (or separate edition thereof) the publisher intends to list as a publication in print.
ISBN stands for International Standard Book Numbering. This is a numeric code that identifies the publisher and each subsequent publication offered by the publisher. ISBN numbers are purchased in blocks of ten or more from RR Bowker, the USA ISBN agency. Currently, a block of 10 ISBN numbers runs $269.95. It is also possible to buy single ISBN numbers through commercial brokers for about $55.00 each. For more information, details can be obtained at http:/www.isbn.org, which is the official website for information and purchase of ISBN numbers through RR Bowker. It is also possible to purchase individual ISBN numbers through broker companies that buy up blocks of the numbers and then sell them individually to self-publishers needing only one number. The web address for such a service is http:/www.isbn4authors.com.
As a self-publisher, it is vital to have your book listed in the Books in Print database maintained by RR Bowker. Once you have assigned an ISBN number to a published work, you can list it with Bowker. This allows book retailers the ability to locate your book and ordering information. Details of this transaction can be obtained on Bowker’s main website at www.Bowker.com. Also important to obtain is an EAN (electronic bar code number) for your book, if you wish to distribute your book through standard book retailers or re-sellers. Also available at www.Bowker.com.
… Assuming you’ve got a manuscript ready to go to print (cover design, illustrations, index, chapter headings, page lay-outs, etc.), your next step is deciding how you want your book to be printed. You can contract with job printers to do the entire thing, many of which are available on the Internet. Or you can invest in some good heavy-duty, high quality laser printers and do it yourself.
The advantage to self-printing is that digital technology means being able to print on demand as orders come in, saving you from having to lay out hundreds or thousands of dollars on pre-printed books that might not sell for quite some time. The disadvantage of self-printing is the time it takes, not to mention the hassles of dealing with temperamental printers, paper jams, toner and drum purchases, etc. And if you print your own books, you also need to find a workable binding system for the books. In any event, unless you contract out for a substantial volume discount with a printing company on production runs over 1000 books, count on printing and binding costs per 100-page book (8.5 inches by 11 inches) of at least seven to ten dollars per book. If you dare to include color illustrations, your cost per unit will be considerably higher.
One way or another, you’ve got your books ready to sell. What do you do about distribution or advertising? No one will buy if there is no ready means of learning about the book’s availability or its desirable features. The bottom line is that advertising is anything but cheap. And as Internet popularity continues to grow, advertising on the World Wide Web gets more and more expensive as demand
rises with it. Advertising is a must, however. Kinetic Energy Press has obtained excellent results in its relationship with the Accident Reconstruction Network as the premier web portal for individuals seeking information about accident-related topics. Banner ads, preferred listing as ARC Book Store Sponsors, and newsletter articles have all proven to be effective means of advertising books published by Kinetic Energy Press and other publishers affiliated with the ARC Network.
It is also important to mention other ways authors can publicize their books. One tried and true method of publicity is providing free copies of a published work to people who regularly write book reviews for various publications. Another good method is purchasing vendor space at trade group meetings where the attending members are likely to be appreciative of the book’s content or theme. Also important is a conscious effort to publicize the book on the various list-serve groups used by people in the profession most closely aligned with the book’s content. And one should not forget email lists, newsletters, and even press releases as valid means of advertising and publicity. Many published experts have also discovered that purchasing copies of their own books from their publishers for the purpose of strategically giving the books away to influential people and clients is an effective marketing strategy. Notable publishing expert Steve Weintraub, publisher and president of Lawyers and Judges Publishing Company, has repeatedly demonstrated how giving books away instead of business cards results in dramatic increases in client volume and repeat retention for published experts who provide consulting services.
There are tremendous opportunities for aspiring authors. There are more than 65,000 magazines and journals published regularly in the USA and Canada. There are also over 48,000 professional associations in the USA, most of which publish a newsletter or magazine. As a result, more than 160,000 people each year in the USA have at least one article or book published. Why can’t you be one of them? If you have a passion for writing, and you really have the desire to get into print, there is no one who can ultimately stop you from doing so. (And it really doesn’t matter how many rejection slips you get in the process!)
We can boil down the process of getting into print into the following list of steps:
•First, know whom you are and what you have to say.
•Second, define your writing idea into an easily communicated and persuasive writing proposal.
•3rd, explore the best and most appropriate choices for publication of your idea.
•4th, do the necessary research to assure your success in your chosen venue.
•5th, fine tune your proposal or project so that you can begin work on doing the writing.
•6th, write, re-write, edit, proofread, and write some more. Bug your friends, colleagues, and associates until you’re certain your writing makes its intended impact.
• Your seventh step involves taking all the extra necessary effort to assure a print-ready manuscript. This includes illustrations, index preparation, necessary releases for quotations or borrowed material, etc.
•8th, either submit your manuscript on schedule to your publisher, or start the process of publishing it yourself.
•9th step is for self-publishers - it involves all the administrative details associated with ISBN, copyrights, and distribution.
•10, enjoy your finished product! You will now have become a published author.
Helpful Resources:
The World Wide Web is a wonderful resource for authors. The following websites are highly recommended for their assistance in the various stages of successful writing and publishing:
•www.Bowker.com
•www.ISBN.org
•www.loc.gov (this is the Library of Congress)
•www.BookAdvertising.net
•www.RJCom.com (printing and publishing help)
•www.selfpublishing.com
•www.SPANnet.org (official website of the Self Publishers Association of North America)
Also recommended are the following books:
•Getting Published - A Guide for Businesspeople And Other Professionals, Gary S. Belkin, Wiley Press, ISBN 0 471 89338-2
•The Complete Guide to Self Publishing, Tom and Marilyn Ross, Writer’s Digest Books, ISBN 0-89879-646-6
Best of luck to you. Hope to see you in print very soon!
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
Rollover testing has typically been conducted with vehicles equipped with outriggers. The outriggers serve two primary purposes: to insure the safety of the test driver and to prevent irreparable damage to the test vehicle. However, outrigger contact usually occurs shortly after two-wheel lift, subsequently affecting the vehicle’s pre-roll trajectory.
Within the last ten years, state-of-the-art technology has allowed for remote control operation of vehicles subject to rollover. However, remote control operation of vehicles subjected to onroad untripped rollovers at highway speeds is costly endeavor. Therefore, very little test data exists, particularly at the point of roll, for vehicles that have experienced steering-induced onroad, untripped rollover.
The purpose of this paper is to present vehicle dynamics data for Sport Utility Vehicles (SUVs) that have been involved in real world, on-road, untripped rollover accidents. This paper compiles and analyzes 34 SUV rollover accidents. The accidents were reconstructed with the pre-roll vehicle motion simulated using the Vehicle Dynamic Analysis Non-Linear (VDANL) computer program. The vehicle dynamic characteristics that were analyzed include: vehicle speed, roll angle, roll rate, yaw angle, yaw rate, lateral acceleration, and sideslip angle. These characteristics were analyzed at two distinct points: two-wheel lift and four-wheel lift. Steer inputs, numbers of rolls, and rollover distances are also presented.
To reduce the occurrence of rollover and the high injury severity associated with these accidents, a better understanding of the vehicle dynamics that lead to rollover important design tools. Furthermore, the data presented in this paper provide valuable information for reconstructing rollover accidents.
A rollover is defined as an event in which the vehicle rotates at least 90 degrees about its longitudinal axis and its tires are no longer in contact with the ground. [1] Rollovers account for 3% of all vehicle accidents involving passenger cars and 8% of all accidents involving SUVs. [2] Although these frequencies are relatively low when compared to all types of vehicle accidents, rollovers are associated with high fatality rates. FARS data for 1991-2000 calendar years show that rollovers accounted for 75% of all fatalities involving SUVs and 42% of all fatalities involving passenger cars. [3]
Typical rollover testing has been conducted with vehicles equipped with outriggers. The outriggers serve two primary purposes: to insure the safety of the test driver and to prevent irreparable damage to the test vehicle. However, outrigger contact usually occurs shortly after two-wheel lift, subsequently affecting the vehicle’s pre-roll trajectory and thus preventing observation of the vehicle dynamic characteristics at point of roll. The point of roll is defined as the point at which there are no longer any vertical tire forces on any of the vehicle’s tires. Within the last ten years, state-of-the-art technology has allowed for remote control operation of vehicles subject to rollover, including braking, accelerating, and steering. However, remote control operation of vehicles subjected to on-road untripped rollovers is a difficult and costly endeavor. Furthermore, most tests that induce rollovers are conducted at speeds no more than 50 miles per hour (mph). Therefore, very little test data exists at the point of roll for vehicles that have experienced steering induced on-road, untripped rollovers, particularly at highway speeds.
(c)2006CrashDataGroupInc
The database presented in this paper is a compilation of 34 real world SUV rollovers. These cases were investigated and reconstructed for the purposes of litigation. The original data base which has been updated and augmented for this research has
Lawrence Wilson Wilson Consulting Daniel Godrick Ian S. Jones & Associates
been compared to FARS data and is representative of SUV single vehicle, first-event rollovers involving fatal and serious injuries. [4] This study was limited to on-road, untripped SUV rollovers where the road grade and super elevations were minimal so that environmental conditions at rollover could be as uniform as possible.
Each accident involved a serious or fatal injury to one or more of the vehicle’s occupants. The investigation for each accident followed a prescribed methodology and included a detailed inspection of the vehicle and accident scene. [5]
Physical evidence on the vehicle related to the rollover was identified, photographed and documented. Physical evidence included: angle and orientation of scrape marks on the vehicle, lateral and vertical roof crush, fluid stains, occupant compartment damage, fender and sheet metal crush and any other notable damage.
The accident scene inspection consisted of identifying and photographing physical evidence that was related to the rollover. Physical evidence included, yaw marks, rim gouges, rim imprints, scrape marks, gouges, glass debris, and miscellaneous debris from the vehicle. Pre-roll yaw marks were established from police measurements, police markings and /or photographs. Reverse projection photogrammetry was used to re-create yaw marks from existing photographs when necessary.
The physical evidence, along with the roadway layout and terrain, was surveyed using a laser transit. Once the vehicle and scene inspections were conducted, this information was combined with existing evidence in order to accurately analyze and reconstruct the rollover accident. For each accident reconstruction the following information was determined: steering inputs and the vehicle’s corresponding pre-roll trajectory, vehicle travel speed prior to loss of control, vehicle speed at point of roll, and rollover trajectory, including number of rolls. Figure 1 is an example of a completed reconstruction.
For each reconstruction, the speed of the vehicle at point of roll was determined by considering the distance from point of roll to final rest. The pre-roll vehicle trajectory was determined by matching the computer output with the documented tire marks leading up to the point of roll. The pre-roll vehicle motion was simulated using the Vehicle Dynamic Analysis Non-Linear (VDANL) computer program. This program is a 3-dimensional vehicle simulation software package that was developed specifically for NHTSA to simulate single vehicle rollover dynamics. [6] Figure 2 shows the simulated vehicle trajectory that corresponds to the accident reconstruction shown in Figure 1. Figures 3a through 3f show the VDANL simulation output.
For more than 30 years, computer simulations have shown the capability to accurately model vehicles involved in limit handling maneuvers. Ford Motor Company detected front wheel-lift in a 1986 prototype Bronco II 4x2 during a simulated ramp steer maneuver of 90 degrees at a speed of 60 mph. [7] Ford incorporated computer simulations into their design process in order to save time on vehicle testing, reduce prototype part costs, and because of the safety concerns typically associated with limit testing.[7] Computer simulations can also be used in the accident reconstruction process to verify steer inputs and speed loss during the pre-roll phase of a rollover accident. In many situations, a basic simulation model such as EDSVS is adequate to provide a reasonable estimate for the steering, braking and speed change of a vehicle during the pre-roll phase of the accident. [8] However, for more precise simulation of at-limit dynamics (i.e. tire loads, lateral acceleration, roll angle, roll rate, and side slip) three-dimensional programs such as VDANL must be used. VDANL has been extensively validated. [6, 9, 10, 11] Figures 4a through 4f show plots of VDANL simulated data compared to vehicle test data for a four-door, four wheel-drive Toyota 4-Runner. The steer maneuver represented in Figure 4a is the ISO 3888 Part 2 Double Lane Change [12] Test data from NHTSA was used as a basis for comparison for the VDANL simulations shown in these figures. [12] The 4-Runner used in the NHTSA tests was model year 2001 with its electronic stability control (ESC) disabled. The 4-Runner used to measure the parameters for the VDANL simulations was model year 1997. Detailed examination of tires, suspension components, and steering components determined that these vehicles were substantially similar. The predicted vehicle response shows extremely close correspondence with the test vehicle results.
To analyze and gain a better understanding of the pre-roll dynamics of SUVs, two specific points on the vehicle’s pre-roll trajectory were chosen: the point at which the vehicle experiences two-wheel lift (TWL) and the point at which the vehicle rolls. The point of roll is defined as the point at which there are no vertical tire forces on any of the tires.
The point at which the vehicle experiences TWL is significant because most drivers do not have the reflexes or skills to control a vehicle experiencing two-wheel lift (TWL) [1]. Consequently, this point is often called the rollover threshold. [1] Since most drivers cannot successfully regain control of a vehicle after TWL, the design of the vehicle as it relates to rollover should be focused on how the vehicle handles up to this point. The dynamic vehicle characteristics at TWL that are presented include: vehicle speed, lateral acceleration, sideslip angle, yaw rate, and chassis roll angle.
“Each accident involved a serious or fatal injury to one or more of the vehicle’s occupants”
These vehicle dynamic characteristics are presented in Figures 5a through 5e, respectively. Table 1 is a summary of the data presented in these figures.
Number of Rolls 3-1/2 1-1/4
Rollover Distance (ft) 188 64.3
Figure 6a shows initial travel speed prior to steer inputs. Figure 6b shows the number of steers in a steering sequence that the driver input that lead to rollover. Figure 6c shows the maximum steer amplitude and Figure 6d shows the maximum steer rate input by the driver for any given steer in a steering sequence that lead to rollover. Table 2 is a summary of the data presented in these figures.
The elapsed time, distance traveled and speed dissipated from TWL to point of roll are shown in Figure 9a through 9c. The elapsed time from TWL to roll is important data considering driver perception-reaction times. The distance traveled and speed dissipated is valuable data for reconstructionists when analyzing the pre-roll yaw marks and determining speed along the pre-roll trajectory. The change in yaw rate from TWL to point of roll is presented in Figure 9d. The yaw rate typically decreases from TWL to point of roll. Table 5 is summary of the data presented in these figures.
Change in Data from Mean Standard Deviation
TWL to Roll
Time (sec) 0.9 0.3
Distance (ft) 75 28
Speed Loss (mph) 8.6 4.5
Yaw Rate -16 16
The point of roll is the point at which four-wheel lift occurs and there are no vertical tire forces. Vehicle dynamic data presented at point of roll includes: vehicle speed, sideslip angle, yaw angle, yaw rate, and roll rate (angular velocity about the vehicle’s longitudinal axis). This data can be compared with rollover conditions of vehicles subject to rollover testing such as the FMVSS 208 Dolly Rollover test in which a vehicle is pitched from a dolly at 30 miles per hour (mph). In this test, the vehicle is oriented at 90 degrees relative to its direction of travel and the vehicle has no yaw velocity. Figures 7a through 7e show the vehicle dynamic data at point of roll. Table 3 is a summary of the data presented in these figures.
Number of rolls and rollover distance are presented in Figures 8a and 8b, respectively. This data is useful to accident reconstructionists because it represents data typical of serious rollover events. Table 4 is summary of the data presented in these figures.
The data presented in the previous section warrant further clarification, discussion and analysis. A more thorough presentation of the data is given in Appendix A. All of the data presented in Figures 5 through 9 involve SUVs that were not equipped with electronic stability control (ESC). ESC is a technological development that has shown promise in maintaining directional and rollover stability. Some vehicles were equipped with anti-lock braking systems (ABS) but in most of the cases drivers did not apply brakes irrespective of whether the vehicle was equipped with ABS.
Figure 9a shows the elapsed time from TWL to point of roll. The mean elapsed time from TWL to point of roll is 0.9 sec. More than eighty percent of the vehicles reached the point of roll less than 1.25 seconds after experiencing TWL. Human factors studies have shown that a typical perception-reaction time is 0.75 seconds for a driver that is alerted to a hazardous situation. [13] Typical reverse steer input times are about 0.5 seconds. [4, 14] Given that there is usually a phase lag of 0.1 to 0.2 seconds between the input of the steer and the reaction of the vehicle, it would take 1.35 to 1.45 sec for a typical driver to identify TWL, react by inputting a corrective steer and have the tires start to react to the corrective steer. This scenario optimistically assumes that a typical driver can sense TWL as soon as it occurs. Thus, this data fully supports Gillespie’s assertion that most drivers do not have the reflexes or skills to control a vehicle experiencing TWL. [1]
The number of driver steer inputs shown in Figure 6b only included reactive emergency avoidance inputs. Steering inputs that characterize the driver’s initial excursion that lead to the emergency steering sequence were not included in this data. Furthermore, this data did not differentiate between a left steer or a right steer. Both directions counted as one steer regardless of direction or amplitude as long as it was a reactive emergency steer maneuver.
Referring to maximum steer amplitude and maximum steer rates shown in Figures 6c and 6d, the largest steer amplitude was typically the last steer in the emergency steer sequence; whereas, steer rates were generally consistent throughout the steer sequence. The minimum steer input was 55 degrees and the maximum input was 229 degrees. These values use a centered steering wheel position as a reference. For example, a 180 degree steer to the left followed by 360 degrees of steering wheel movement to the right in which the steering wheel stops moving 180 degrees clockwise from its centered position is defined as a two-steer input with a maximum steer amplitude of 180 degrees. The minimum steer rate was 262 deg/s and the maximum steer rate was 802 deg/s with a mean steering rate of 590 deg/s.
Jones, et al. reported similar results in a related study [4]. In the Jones study, research conducted by Nissan Motor Corporation and NHTSA was summarized as a means for comparison. The steering profiles i.e. amplitude and rate from these sources were quite similar. [4, 15, 16]. In a more recent study, Honda Motor Corporation used steer amplitudes of 90 deg to 200 deg and steer rates of 720 deg/s when conducting rollover stability tests intended to simulate severe emergency avoidance maneuvers. [14] When using the Road Edge Recovery maneuver to evaluate the dynamic rollover stability of SUVs, NHTSA used maximum steer amplitudes ranging from 230 to 312 degrees, and a steering rate of 720 deg/s. [17]
The steer amplitudes and steer rates implemented in the Honda and NHTSA studies were greater than the values presented in this paper. This is to be expected because the NHTSA and Honda studies implemented steering inputs that were intended to represent the high end of the range of typical emergency steer maneuvers.
Initial travel speed presented in Figure 6a and Table 2 show that the mean travel speed is 65 mph with a minimum initial speed of 44 mph and a maximum initial speed of 85 mph. By comparison, the Honda and NHTSA studies were conducted at speeds ranging from 35 mph to 57 mph. [14, 17] These tests do not model the travel speeds presented in this paper. However, it could be argued, that the increased severity of the steers implemented by these tests, combined with the sensitivity of vehicle handling to steering dwell time, as cited by Honda, justify the comparably lower speeds at which these test are conducted.
Comparing the data in Tables 1, 2, and 3 shows the mean initial speed is 64 mph, the mean speed at TWL is 58 mph and the mean speed at roll is 50 mph. Thus, most of the speed is dissipated from TWL to point of roll even though most pre-roll trajectories typically involved two or three steers. Recalling that most of the pre-roll trajectories did not involve braking and comparing side
slip values in Figures 5c and 7b, most of the speed during the preroll trajectory was dissipated as a consequence of increased side slip from TWL to roll.
Referring to Figure 5b, the minimum and maximum values of lateral acceleration at TWL are 0.59 g and 0.94 g, respectively. NHTSA reported slide friction values ranging from 0.82 to 0.89 and peak friction values of 0.92 to 1.01 for the test surfaces used to conduct rollover stability testing. [17] It is important to recognize that peak friction values need not be attained in order for TWL to occur. Another characteristic to be considered is static stability factor (SSF). All cases presented in this paper involved SUVs with SSFs less than 1.15. Furthermore, the lowest lateral acceleration to cause TWL was 0.59 g and involved a fully loaded SUV with a SSF below 1.0. This data represents SUVs from model year 1984 though model year 2001. Since the advent of NHTSA’s Five-Star Rating System more recent model year SUVs have improved SSFs and improved suspensions. [18] Therefore, the authors expect that the mean average lateral acceleration at TWL would increase if later model year SUVs were included.
Side slip is defined as the difference between the direction that the vehicle is oriented and the direction that the vehicle is traveling. More than ninety percent of the vehicles in Figure 5c experienced TWL at side slip angles less than 15 degrees resulting in a tight distribution of data skewed toward the lower values of side slip. Conversely, Figure 7b shows a very even distribution of side slip angles at point of roll. All of the SUVs that experienced rollover at side slip angles less than 10 degrees were circa 1980 – 1989 and had SSFs of less than 1.05. This implies that these SUVs experienced very little slide out prior to rolling. SUVs that have higher SSFs and improved suspension components typically experience higher side slip prior to TWL and rollover. Higher side slip angles result in higher lateral accelerations and these lateral accelerations are sustained for a longer period of time, thus facilitating rollover. Many SUVs introduced within the last three years, sometimes referred to as hybrid SUVs are built on passenger car platforms and have SSFs greater than 1.2. [18] These SUVs should not experience steering-induced, on-road, untripped rollovers providing the roll stiffness distribution and roll gradient are within acceptable limits.
Because yaw and side slip are related, it is expected that yaw at point of roll is affected by the vehicle’s SSF. Referring to Figure 7c, only one case rolled at a yaw angle of 20 degrees or less. The vehicle involved in this case had a SSF of less than 1.0. Similar to side slip, the yaw at point of roll, shown in Figure 7c has an even distribution but gives a well-defined range. Ninety percent of all cases were yawed between 30 and 70 degrees at point of roll. For rollover reconstruction, yaw and side slip at roll can be verified by examination of the yaw marks at point of roll, scrapes and gouges during first touchdown, and orientation of scrape marks on the vehicle. The yaw marks and location of scrapes/gouges relative to the yaw marks can provide orientation of the vehicle at point of roll and direction of travel. The scrape marks on the vehicle provide a means of cross-checking the yaw and side slip at roll. The scrape marks that represent first touch-
down on the vehicle should be equal to or greater than the side slip angle at point of roll since the vehicle continues to yaw in the same direction from point of roll to first touchdown.
Comparing the average (mean) values for side slip and yaw in Table 3 shows that side slip at point of roll is, on average, 17 degrees less than yaw at point of roll. This is expected because none of these cases occurred with fully locked brakes, a condition that would allow side slip and yaw to be equivalent. Because the vehicle’s tires are still rolling up to the point of roll, side slip should be less than yaw.
Yaw rate typically decreases from TWL to point of roll. Table 3 shows the average reduction in yaw rate from TWL to roll is 16 deg/s. Figure 9d shows that all but 4 cases, accounting for almost ninety percent of the represented population, experienced a reduction in yaw rate from TWL to roll. A possible explanation of this somewhat counter-intuitive trend is that a steering effect in the opposite direction of yaw is generated when the vehicle rolls onto the sides of the two tires that are still on the pavement or steering force is simply reduced when only two wheels are on the ground.
Several cases were excluded from this data because the rollover was precipitated by a collision and therefore were not classified as first event, single-vehicle rollovers. As a side note for accident reconstructionists, these cases generally exhibited an increase in yaw rate from TWL to point of roll.
The minimum and maximum values for roll rate at point of roll shown in Figure 7e are 150 deg/s and 353 deg/s, respectively and about two-thirds of the cases presented in this figure have roll rates ranging from 250 deg/s to 350 deg/s. In accident reconstructions, the roll rate at point of roll can be compared with the average roll rate during the first roll to make sure that the values are compatible. Average roll rate during the first roll can be determined from physical evidence if the speed at point of roll and distance of the first roll are known. Vehicle speed at point of roll can be calculated if the distance from point of roll to final rest is known. [5]
Roll rates can increase during the rollover trajectory. Increased roll rate usually occurs when translational velocity decreases and rotational velocity is increases. The physical evidence associated with increased roll rate include the vehicle yawed close to 90 degrees relative to its direction of travel and heavy rim gouges at the start of the roll which load the suspension causing more rebound off the ground. These heavy rim gouges usually have corresponding damage to the rims that caused the gouge marks. Also, a roll that has an increased roll rate within a rollover sequence usually exhibits lighter scrape marks on the vehicle, assuming all of the rolls were on the same surface. One exception is if some rolls occur on pavement while others occur on grass because scrapes left on the vehicle from pavement are typically more severe than scrape marks caused by grass.
Referring to Table 4, dividing the average rollover distance (188 feet) by the average number of rolls (3.46) yields an average distance per roll of 54 feet. The first roll is often the longest roll in the rollover trajectory thereby exhibiting a greater than average distance per roll. The first roll typically has a longitudinal orientation relative to its direction of travel, making the vehicle more prone to sliding and less prone to rolling thereby extending the distance traveled during this roll. The high vehicle speed during the first roll also contributes to typically making it the longest roll in the rollover sequence. A longer than average roll distance during the first roll is usually accompanied by long, heavy scrape marks on the vehicle biased towards the front of the vehicle. The bias of scrapes toward the front is a consequence of its forward, longitudinal orientation compared to other rolls in the trajectory. The heavy nature of the scrape marks is a consequence of the vehicle sliding more and consequently, having longer ground-contact time.
As the vehicle slows during subsequent rolls and approaches a 90 degree orientation relative to its direction of travel, the distance per roll decreases. As mentioned in the previous sub-section, this effect is enhanced if the roll rate is increased by heavy rim contact at the start of a roll. Increased roll rate and a decreased distance per roll is usually accompanied by lighter and shorter scrape marks that are more evenly distributed, front to rear, along the vehicle.
The data presented in this paper is a reliable and valuable source of information to automotive engineers interested in rollover testing of SUVs because it provides a basis for realistic conditions, including driver inputs, that lead to rollover and a realistic basis for conditions at the point of roll. For accident reconstructionists attempting to reconstruct rollovers involving SUVs, this data provides a sound basis from which their results can be compared.
The data presented in this paper is based on accident reconstructions and VDANL simulations that were conducted by Ian S. Jones & Associates. The authors would like to thank Dr. Ian S. Jones for the use of this valuable data, without which this paper would not have been possible.
1.Gillespie, Thomas D., Fundamentals of Vehicle Dynamics, Society of Automotive Engineers, Inc.
2.Subramanian, Rajesh, “Analysis of Crashes Involving 15-Passenger Vans”, DOT HS 809 735, May, 2004.
3.“Evaluation of the Rollover Propensity of 15-passenger Vans”, Safety Report NTSB/SR-02/03, National Transportation Safety Board, October 15, 2002.
4.Jones, I. S., L. A. Wilson, and R. A. Whitfield, “Emergency steering in fatal and serious injury sport utility vehicle rollover crashes”, IMechE Paper C567/044,2000, 2000.
5.Jones, Ian S., Wilson, Lawrence A., “Techniques for the Reconstruction of Rollover Accidents Involving Sport Utility Vehicles, Light Trucks and Minivans”, SAE paper 2000-01-0851, 2000.
6.Allen, R. W., H. T. Szostak, T. J. Rosenthal, D. H. Klyde, and K. J. Owens, “Vehicle Dynamic Stability and Rollover”, NHTSA/
The Accident Reconstruction Calculator™ provides you with a complete suite of equations.
Momentum
Time-Distance
Basic Motion
Low-Speed
Energy
Force
Delta-V
Crush
Auto-Pedestrian ProjectileMotion andmore…
Theequationsyougrewupwith.Taketheguessworkoutofyourworkwith this user-friendly program that has benefited over 500 police agencies and accident reconstruction firms throughout the United States. Visit our website forcurrentpriceandorderform.Whilebrowsingourwebsite,takealookat othercollisionreconstructionproductsandourtrainingclassesaswell.
WeCARE Products
228PleasantStreetRoseville,CA95678 Bus.(916)782-4654Fax(916)782-4878
WeCARE@rudydegger.comhttp://www.rudydegger.com/WeCARE
DOT HS 807 956, June, 1992.
7.Antoun, R.J., P. B. Hackert, M. C. O’Leary, A. Sitchin, “Vehicle Dynamic Handling Computer Simulation – Model Development, Correlation, and Application Using ADAMS”, SAE Paper 860574, 1986.
8.Martinez, J. Ed and Richard J. Schlueter, “A Primer on the Reconstruction and Presentation of Rollover Accidents”, SAE Paper 960647, 1996.
9.Allen, R. Wade, Dong-chan Lee, David H. Klyde, et al., “Vehicle Dynamics Validation for Real-Time Simulation, Systems Technology, Inc., October 19, 2001.
10.Chrstos, J. P., and G. J. Heydinger, “Evaluation of NHTSA Light Vehicle Handling Simulations, NHTSA Final Report No. DOTHS-807-868, 1992.
11.Chrstos, J. P. and G. J. Heydinger, “Evaluation of VDANL and VDM RoAD for Predicting the Vehicle Dynamics of a 1994 Ford Taurus, SAE Paper 2000-01-1620, May, 2000.
12.Forkenbrock, Garrick J., et al., “A Comprehensive Evaluation of Test Maneuvers That May Induce On-Road, Untripped Light Vehicle Rollover - Phase IV of NHTSA’s Light Vehicle Rollover Research Program”, DOT-HS-809-513, October, 2002.
13.Olsen, Paul, Forensic Aspects of Driver Perception and Response, Lawyers and Judges Publishing Company, 1996.
14.Kebschull, Brian, Keisuke Ishii, Mark Ernst, “Rollover Resistance Test Procedure Involving Maximum Roll Momentum”, Proceedings of the 18th International Technical Conference on the Enhanced Safety of Vehicles, Nagoya, Japan, 2003.
15.Maeda, Terou, Namio Irie, et al., “Performance of Driver-Vehicle System in Emergency Avoidance”, SAE Paper 770130, 1977.
16.Mazzae, Elizabeth N., Frank Barickman, et al. “Driver Crash Avoidance Behavior with ABS in an Intersection Incursion Scenario on Dry Versus Wet Pavement”, SAE Paper 1999-01-1288, 1999.
17.Forkenbrock, Garrick J., et al., “An Experimental Evaluation of 26 Light Vehicles Using Test Maneuvers That May Induce On-Road, Untripped Rollover and a Discussion of NHTSA’s Refined Test Procedures – Phases VI and VII of NHTSA’s Light Vehicle Rollover Research Program”, DOT-HS-809-547, October, 2003.
18.Walz, Marie, “Trends in the Static Stability Factor of Passenger Cars, Light Trucks, and Vans”, DOT-HS-809-868, June, 2005.
CONTACT
Lawrence Wilson can be contacted via e-mail at wilsonconsulting@verizon.net.
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
(c)2006CrashDataGroupInc
Let’s get this out of the way up front: yes, we pre-cut the target car so it would split in half. So what? Let’s keep this in perspective: this case study is about momentum ... particle momentum ... tracking masses moving in specific directions at specific speeds ... vectors. We are not applying any sort of energy analysis ... no crush damage in this analysis. Maybe, if we had some idea of specifically what impact (pun intended) the pre-test “dotted line” had on the stiffness of the target car in this example, if we could quantify the change we made in the structural integrity of the target car, we might entertain that as a separate intellectual exercise, but that was not the plan here so before we get sidetracked with inane permutations, too mindless to consider, let’s stay on track: this is a momentum exercise; complete, simple, independent of energy1.
On June 5th, 2006 during the ARC-CSI Conference in Las Vegas, a crash test was conducted using a 2000 Chevrolet Malibu as the designated bullet vehicle and a 1991 Geo Metro at the target vehicle. It was the Geo that, prior to the test, was pre-cut to ensure that during the crash the Geo would separate into two pieces; two pieces of known dimensions.
Since this was crash test - it was planned and conducted - there are facts about the crash that an investigator who would be working this crash after the fact would not have normally
known with the same degree of certainty as we have here (that’s why crash tests make good case problems!). In order to attempt to make this analysis as realistic as possible, let’s approach it as if this crash had occurred on a public road, with the analysis, when possible, based solely on the scene evidence. There were facts known about the movement of the vehicles post collision that will be ignored in this case problem, since they would not have been known to a investigator working the collision in “the real world,” but that’s the only meaningful difference here.
At this point, the trained investigators and attentive readers will realize that the crash data from the June 5 ARC-CSI Conference in Las Vegas is right here before you, in your hands at this very moment: it’s on the DVD which is part of this issue of Collision. So,
you could simply go to that DVD and review the video and analysis from the conference but if you did that, you wouldn’t be taking advantage of this as a case study so do that later, OK?
You are faced with a crash with little physical evidence left afterward (in particular limited roadway marks), a bullet vehicle which is CDR System accessible, and a target car that’s been split in half in the crash. What do you do? Try to figure out a way to claim some information and effectively analyze the wreck? Or invoke possibly the most meaningless assumption: that being when a car is split in half (or if there’s a secondary contact, or someone moved a car post impact or whatever the inane, perverse topical myth) that you “just can’t ‘do’ momentum?” Well, let’s see what YOU do...
This test involves a 2000 Chevy Malibu (3039lbs / 1379kg) striking a 1991 Geo Metro (1625 lbs / 737kg). After the crash, a download of the Malibu airbag control module’s event data recorder component shows the delta-V for the Malibu to have been -6.3mph (on the X axis) in a non-deployment data file. That data also showed the last pre-crash speed observed to have been 36 mph.
The curb weight for the Malibu was 3039 pounds, while the Geo’s curb weight was 1625 pounds, with approximately 61% resting on the front wheels and 39% on the rear wheels. In the momentum calculations the weight of the Malibu was increased by 225 pounds to account for the driver’s weight and any equipment that had been installed in the vehicle for the crash test. As is the case of most real word cases this weight
was an approximation of the additional weight in the vehicle since the driver and other equipment was not physically weighed.
The front and rear weights to the Geo were adjusted to account for the location of the split. Since the Geo split behind the center mass, the front and rear portions would not have maintained the 61/39 percent ratio that existed prior to the collision. The rear portion of the Geo accounted for approximately 25% of the vehicle’s pre collision weight while the front section made up approximately 75% of the total pre collision weight. These estimations resulted in weight for the rear portion of the vehicle of 406 pounds and 1219 pounds for the front portion.
Despite the otherwise apparent violent nature of this collision and similar movement of the vehicles when observed on the video, as in many real word collisions, there was little post impact scene evidence available making the reconstruction of this collision - again, like many real world crashes - difficult. There were no gouge marks in the pavement that could have been used to help identify the area of impact. Nor were there many tire or scrape marks that would have been helpful in determining post impact movement of either portion of the Geo with much specificity. However, there were two tire scuffs produced in the area of impact. The right rear wheel of the Geo produced the first mark while the right front wheel of the Malibu produced the second mark.
Now, the question for this case problem is....how fast are the Geo and Malibu moving at impact? (Yes, they were both moving at impact.) You’d have to evaluate the scene (the scene maps are on the attached DVD in multiple formats) to find impact and separation angles, you’ll have to sort out effective roadway friction for the involved vehicles or parts and of course, post impact distances. Once you’ve given that some thought, turn to page 102 for one evaluation, there may be others. Or, you could take the inept position to merely adopt the CDR recovered data from the Malibu and assume it’s at 36mph...or not!
1.In case you missed the discussion of the pre-test cutting... Even though some may describe this pre-test cutting as “cheating” as it relates to the separation of the Geo, the fact that the vehicle was pre-cut does nothing to change Newton’s Laws and their application to collision reconstruction. For this analysis, it makes no difference how much, or how little energy was required to separate a vehicle into two pieces, conservation of momentum still applies (conservation of momentum and energy are not synonymous concepts). So, the speeds of the vehicle may still be calculated using traditional collision reconstruction methods based on conservation of momentum.
There’s not much damage to this Malibu for it having split a car in half...why not?
All SDM recorded data is measured, calculated, and stored internally, except for the following:
-Vehicle Speed, Engine Speed, and Percent Throttle data are transmitted once a second by the Powertrain Control Module (PCM), via the vehicle’s communication network, to the SDM.
-Brake Switch Circuit Status data is transmitted once a second by either the ABS module or the PCM, via the vehicle’s communication network, to the SDM.
-The SDM may obtain Belt Switch Circuit Status data a number of different ways, depending on the vehicle architecture. Some switches are wired directly to the SDM, while others may obtain the data from various vehicle control modules, via the vehicle’ s communication network.
-The Passenger Front Air Bag Suppression Switch Circuit is wired directly to the SDM.
Recorded Crash Events: There are two types of SDM recorded crash events. The first is the Non-Deployment Event. A Non-Deployment Event is an event severe enough to “wake up” the sensing algorithm but not severe enough to deploy the air bag(s). It contains Pre-Crash and Crash data. The SDM can store up to one Non-Deployment Event. This event may be overwritten by another NonDeployment Event. This event will be cleared by the SDM after the ignition has been cycled 250 times. The second type of SDM recorded crash event is the Deployment Event. It also contains Pre-Crash and Crash data. The SDM can store up to two different Deployment Events, if they occur within five seconds of one another. Deployment Events cannot be overwritten or cleared from the SDM. Once the SDM has deployed the air bag, the SDM must be replaced. The data in the Non-Deployment Event file will be locked after a Deployment Event, if the Non-Deployment Event occurred within 5 seconds before the Deployment Event unless a Deployment Level Event occurs within 5 seconds after the Deployment Event, and then the Deployment Level Event will overwrite the Non-Deployment Event file.
SDM Data Limitations: -SDM Adjusted Algorithm Forward Velocity Change: Once the crash data is downloaded, the CDR tool mathematically adjusts the recorded algorithm forward velocity data to generate an adjusted algorithm forward velocity change that may more closely approximate the forward velocity change the sensing system experienced during the recorded portion of the event. The adjustment takes place within the downloading tool and does not affect the crash data, which remains stored in the SDM. The SDM Adjusted Algorithm Forward Velocity Change may not closely approximate what the sensing system experienced in all types of events. For example, if a crash is preceded by other common events, such as rough road, struck objects, or off-road travel, the SDM Adjusted Algorithm Forward Velocity Change may be less than and some times significantly less than the actual forward velocity change the sensing system experienced. This data should be examined in conjunction with other available physical evidence from the vehicle and scene when assessing occupant or vehicle forward velocity change. For Deployment Events and Deployment Level Events, the SDM will record 100 milliseconds of data after deployment criteria is met and up to 50 milliseconds before deployment criteria is met. The maximum value that can be recorded for SDM Adjusted Algorithm Forward Velocity Change is about 112 MPH.
-SDM Recorded Vehicle Speed accuracy can be affected if the vehicle has had the tire size or the final drive axle ratio changed from the factory build specifications.
on:
-Brake Switch Circuit Status indicates the status of the brake switch circuit.
-Some of the Pre-Crash data may be recorded after Algorithm Enable (AE). This may happen in situations involving relatively "soft" crash pulses or those that take place over a relatively longer period of time. If this occurs, it may affect the reported precrash data values, but does not affect other data such as SDM Adjusted Algorithm Forward Velocity Change.
-Pre-Crash Electronic Data Validity Che ck Status indicates “Data Invalid” if t he SDM receive an invalid message from the module sending the pre-crash data.
-Driver’s Belt Switch Circuit Status indicates the status of the driver’s seat belt switch circuit. If the vehicle’s electrica l system is compromised during a crash, the state of the Driver’s Belt Switch Circuit ma y be reported other than the actual state.
-Passenger Front Air Bag Suppression Switch Circuit Status indicates the status of the suppression switch circuit.
-The Time Between Events is displayed in seconds. If the time between the two events is greater than five seconds, “N/A” is displayed in place of the time.
-If power to the SDM is lost during a crash event, all or part of the crash record may not be recorded.
-If the vehicle is a 20002002 Chevrolet Cavalier Z24 or a Pontiac Sunfire GT, with a manual transmission (RPO MM5) and a 2.4L engine (RPO LD9), the Brake Switch Circuit Status data will be reported in the opposite state than what actually occurred, e.g. an actual brake switch status of “ON” will be reported as “OFF”.
SDM Data Source: 1G1ND52JXY6226318
on:
Page 1 of 4
It is not uncommon for a slow-moving passenger vehicle (PV) to be struck in the side by a commercial motor vehicle (CMV). When this occurs, the PV often remains engaged with the CMV to their final rest position. Evaluating an appropriate post-impact drag factor requires assessing the frictional contribution of both vehicles. Two crash tests were conducted in which a stationary PV was struck by a CMV traveling at approximately 40 mph (64 kph) in order to assess the overall post-impact drag factor. It was found that the combined units slowed at a rate higher than the CMV alone, but lower than the skidding value of the passenger vehicle alone, and commensurate with the mass-ratio of the two vehicles involved. Standard crash analysis techniques were found to accurately predict the CMV’s pre-crash speed.
Two collisions were staged in which a heavy commercial motor vehicle (CMV) impacted a stationary passenger car on one side. The first test (Crash #1) involved a bobtail International cabover truck-tractor striking a Chevrolet Lumina, while the second test (Crash #2) involved a GMC conventional truck-tractor pulling a lightly loaded flat-bed trailer which struck a Chevrolet S-10 Blazer. For both tests, the CMV operator initiated an emergency stop immediately prior to the impact.
The roadway coefficient of friction at the test area was determined by conducting locked-wheel skid tests with three passenger vehicles. Test skids were also conducted using both CMVs to determine their emergency-stop deceleration rates. Neither CMV was equipped with an anti-lock brake system (ABS), consequently, most of the wheels locked during an emergency stop.
Additionally, the CMV deceleration rates were estimated using the technique promulgated by Heusser [1992], based on vehicle parameters and measured brake adjustments.
The overall post-impact deceleration rate was evaluated based on both VC3000 and Stalker data, and compared to a standard weight-ratio-based momentum analysis.
The test area was a concrete apron at Cecil Field in Jacksonville, Florida. Skid tests were conducted approximately one week prior to the crash test, which occurred on April 25, 2006, as part of IPTM’s 2006 Special Problems seminar. For all tests and crashes, the weather was warm and dry.
Locked wheel emergency-stop skid tests were conducted between 35 and 43 miles per hour (57 and 69 kph) using three vehicles: a 1999 Chevrolet Astro van bearing the vehicle identification number (VIN) 1GNDM19W0XB171453, a 2006 Hyundai Sonata 4-door sedan bearing VIN 5NPEU46FX6H091142, and a 2006 Chevrolet Cobalt 4-door sedan bearing VIN 1G1AL55F467759992. The Astro’s rear wheels could not be reliably locked by the service brake, so tests conducted with only the service brake pedal were discarded. The reported value was the result of simultaneous application of the service brake pedal and the emergency brake pedal, which resulted in four-wheel lockup, and a slightly higher braking rate.
The average skidding acceleration of each vehicle was evaluated using a Vericom VC3000 accelerometer and a Stalker ATS radar unit. The acceleration recorded by the VC3000 was averaged between the point where acceleration reached -0.2 g’s and the point where the vehicle was stopped. The Stalker data was evaluated as the change in speed divided by the change in time, starting from the point where the calculated acceleration rate approached -0.2g’s to the point where the radar unit ceased recording the vehicle’s speed, which was typically at 3 to 5 mph. Consistent with other research (Bartlett et al, 2006), the average acceleration recorded during the five locked-wheel skid tests as measured by the two techniques were essentially the same, and are shown in Table 1.
The first target vehicle was 1997 Chevrolet Lumina LS 4-door sedan which had an overall length of 201 inches (510.5 cm), a wheelbase of 108 inches (274.3 cm), and an overall width of 72 inches (182.9 cm). It weighed 3250 pounds (1474.2 kg). The Lumina’s basic dimensions and weights are shown in Figure 1.
The target vehicle for the second crash test was a Chevrolet Blazer LT which had an overall length of 177 inches (449.6 cm), a 107 inch (271.8 cm) wheelbase, and an overall width of 65 in (165.1 cm). The vehicle’s weight was measured to be 3450 lbs (1565 kg). The Blazer’s basic dimensions and weights are shown in Figure 2.
The 1991 International model 9700 6x4 cab-over trucktractor had an overall length of 250.5 inches (6.36 m), a 144 inch (3.66 m) wheelbase, and an overall width of 96 inches (2.43 m). The unit was weighed by the Florida Department of Transportation’s Commercial Motor Vehicle Enforcement Unit
and had a total weight, including existing fluids and weight of the driver, of 16,550 lbs (7,507 kg). The general dimensions and axle weights are shown in Figure 3.
Due to the very low weight carried by the drive axles, even modest brake applications resulted in almost immediate rear-wheel lockup with attendant axle-tramp. This behavior was correctly predicted by the brake force analysis using a truck-tire skid value of 80% of the measured average PV skid value, and an application air pressure of 90psi. The measured brake parameters are shown in Table 2.The Heusser-technique brake force analysis predicted a skidding acceleration of -0.59g’s for the International.
Vertical Slack Adj. Push Rod Drum Tire Rolling
Four emergency skid-tests were conducted with entry speeds between 43.5 and 43.8 mph (70 kph). As with the passenger vehicles, the speeds were measured with a Stalker RADAR unit and calculated from VC3000 data. During the first skid test, the tractor rotated nearly 180 degrees during the stop, resulting in locked front wheels over some of the travel. This test was discarded as not representing a skidding event. On the subsequent three tests, the operator more effectively kept the vehicle pointed in the direction of travel, but some cab rotation did occur. This resulted in the VC3000 reporting lower-thantrue calculated entry speeds and higher-than-true average accelerations for all three tests. The Stalker data was not susceptible to this error, however, and those results were -0.565, -0.546, and -0.575 g’s with an average emergency braking rate of -0.56 ± 0.015 g’s, which compares well with the predicted value.
The 1992 White-GMC Model WG42T truck-tractor was towing a 2000 Fontaine drop-deck semi-trailer carrying 6,000 pounds (2,721 kg) of steel weights chained in place near the middle of the deck. The tractor and trailer combination had an overall length of 738 inches, a tractor wheelbase of 144 inches and an overall width of 96 inches. The Florida Department of Transportation’s Commercial Motor Vehicle Enforcement Unit measured the combinations’ weight to be 39,600 pounds (17,962 kg) including existing fluids and the driver. Figure 4 shows the general dimensions and weights for the GMC and Fontaine.
The GMC’s brake system was documented, with measurements shown in Figure 3. The anticipated braking performance for the
combination was calculated to be -0.58 g’s based on the Heusser technique.
Four skid tests were conducted with the GMC combination. It stopped straight on all four occasions. The speeds at the start of braking as recorded by the Stalker ranged from 39.6 to 40.4 mph (63.5-64.8 kph). The average accelerations measured with the Stalker were -0.562 ± 0.028 g’s, while the VC3000 (mounted centrally on the windshield) reported -0.585 ± 0.023 g’s. Both values compare well with the predicted value of -0.58 g’s.
Crash Test #1: International versus Lumina
The International truck-tractor struck the Lumina slightly behind its center of mass on the passenger side. The two vehicles remained engaged for 43 feet to final rest, with the Lumina rotating clockwise approximately 39 degrees and the International rotating approximately 8 degrees, as shown in Figure 5
The entry speed at the start of braking 48 feet prior to first contact was measured to be 43 mph, with an impact speed of 35 mph, and a post impact speed of 29 mph, as shown in Figure 6. The post-impact deceleration rate was -0.65 to -0.69 g’s depending on method of evaluation (Stalker, VC3000, Delta V/Delta -T, or stop distance). The Stalker data was collected in the lightest smoothing mode. The VC data shown in Figure 6 was smoothed with a forward-backward 10-boxcar average.
(c)2006CrashDataGroupInc
Based on the truck speed at impact, the principal of conservation of linear momentum predicts a post-impact speed of 29.2 mph, which is consistent with the measured data. The post-impact drag factor can be estimated based on the skid-test results and the vehicle weights as follows:
(Weight of Truck)(*fTruck) + (Weight of Car)(*fCar) = ftotal Weight of Both
(16,550 lb)*(-0.56) + (3,250 lb)*(-0.788) = -0.60 g’s
(16,550 + 3,250) lb
This is somewhat lower than that measured in the test. It is hypothesized that dynamic weight transfer and inter-vehicle weight
transfer are primarily responsible for the discrepancy. During the crash, the International’s front bumper over-rode the Lumina’s doorsill and transferred enough weight onto the Lumina that the International’s front right wheel was locked during the post-impact slide.
Crash Test #2: GMC Tractor-trailer versus Blazer
The GMC model truck-tractor pulling the double-drop trailer struck the Blazer between the axles on its right (passenger) side, with the GMC’s front bumper just over-riding the Blazer’s right frame rail, mechanically engaging the Blazer chassis for the duration of the post-impact travel. Figure 7 shows the post-impact scene diagram
Figure 8 shows the speed and acceleration data collected with the Stalker and VC3000 during the impact. The speed at brake application was 41.4 mph (66.4 kph), approximately 18 feet (5.5 m) prior to impact. The truck slowed to an impact speed of 35.9 mph (57.6 kph). Post impact speed was 32 mph (51 kph). This agrees reasonably well with the momentum-based expectation of 33.0 mph (52.9 kph). The two vehicles traveled 55 feet (16.8 m) to final rest.
The experimentally determined drag factor for the entire post collision slide was found to be 0.66 based on Stalker RADAR data, 0.69 based on the Vericom, 0.62 based on post impact travel distance and measured impact speed
Unlike the results from Crash Test #1, these rates are closer to the truck based drag factor than the Blazer skidding alone, as a result of the much higher mass ratio (11.5 to 1 versus 5.1 to 1 for the first crash test). The mass-ratio-predicted deceleration was only -.60 g’s, which is less than actual. It is hypothesized that weight transfer due to braking and inter-vehicle weight transfer due to the mechanical engagement of the Blazer on the front of the GMC is responsible for this discrepancy.
1)The skidding drag factor for a bobtail tractor and a lightly loaded tractor trailer, both with brakes in good repair, were found to be approximately 70% of the skidding drag factor of three passenger vehicles.
2)Brake force calculations using the Heusser approach were reasonably accurate.
3) In both tests, the post collision drag factor of the CMV-PV combination was between the drag factor of the two vehicles individually, with the drag factor of the heavier vehicle dominating.
4)Dynamic weight shift and inter-vehicle load transfer during the impact caused the actual post-impact drag factor in both tests to be higher than predicted using static axle weights.
Heusser, Ron, Heavy Truck Deceleration Rate as a Function of Brake Adjustment, SAE paper 910126 Bartlett, Wade, et al, Comparison of Drag Sled and Skidding Vehicle Drag Factors on Dry Roadways, SAE Paper. 2006-01-1398
Want to explore this topic further?
At the 2007 ARC-CSI Conference in Las Vegas, two of the many instrumented and documented crash tests conducted involved an air brake heavy vehicle (a fire engine no less!) striking a car in one and then a van in another. So, where can you find that crash data for review and further examination? Check out the Conference DVD with this issue of Collision!
Considering now you’ve gone through the diagrams and looked over the data and photos on the DVD (maybe even checked out the video which would, of course give you an edge), we might contemplate this analysis starting with evaluating the post-crash scene evidence:
In the collision scene, there was a faint tire scuff produced by the right front wheel of the Malibu. This mark measured approximately 42 feet in length. After the initial 5.5feet of travel this mark veered to the right changing direction by approximately 22deg, indicating that there was some type of steering input, whether driver or collision induced, causing the Malibu to change direction after it had separated from the Metro. An inspection of the Malibu after the collision did establish the right front wheel sustained superficial damage. Nonetheless, the wheel was not locked, was not noticeably out of alignment, and the Malibu remained driveable after the collision. For the sake of this case study, you may adopt that the coefficient of friction for the road surface is 0.80. There were two tire scuffs produced in the area of impact. The right rear wheel of the Geo produced the first mark while the right front wheel of the Malibu produced the second mark.
Post impact velocity calculations are arguably the area within the larger model that requires the most speculation on behalf of the investigator, and is the area
where values used in the calculations may be the hardest to defend if confronted.
The distance used to determine the post impact velocities for both vehicles was measured from the approximate center mass of each vehicle at the area of maximum engagement to their respective final resting positions, not from the area of impact to final rest. Admittedly, there would be little difference in the final calculations if a post impact distance from impact to final rest, which is the normal practice of many investigators, was used in the calculations verses the slightly shorter distance from maximum engagement/separation. Nonetheless, the post impact velocities of the involved vehicles are not defined until maximum engagement has occurred. If the post impact velocities were defined at impact, then there would be no delta-t for the collision. Instead the vehicles would instantly go from their impact velocity to their respective post impact velocities, which simply is not the case.
Since establishing the area of maximum engagement is a somewhat subjective process, especially in this case where there was little physical evidence that could be used to identify its location, the distances used in the subsequent calculations could easy vary a short distance. The post impact distance identified for the Malibu was approximately 54.5 feet. The post impact distance for the front portion of the Geo was determined to be approximately 4.5 feet while the distance traveled by the rear portion of the Geo was determined to be approximately 48.5 feet.
The coefficient of friction for the road surface has been given for this case problem as 0.80. This was based on various testing conducted throughout the day at locations throughout the testing site. In a larger sense, even though the coefficient of friction of road surfaces appears to be an area within the field of collision reconstruction that is often extensively debated, its actual applied value is not overly sensitive in the larger model. If the value of the coefficient of friction was ranged from a 0.75 to 0.85 in this case, it will only cause the impact velocity of Malibu to change approximately +/-1 mph, while the impact speed of the Geo will range approximately +/-0.5 mph.
Ironically, often one of the least debated values used in any individual reconstruction is the percentage of braking that a particular vehicle achieved post collision although this is the one single value that may contain the greatest variation, be the hardest to defend or support through the use of physical evidence, and may require the investigator to draw on past experiences that cannot be easily articulated. Unless a vehicle has locked all four wheels, which is not common, it cannot be assumed that the vehicle is slowing at 100% effective braking (slowing) post collision. Nonetheless, we’ve all too often seen analysts simply adopt the 100% value rather than reason through the dynamics of the vehicle(s) what the relative percentages should actually be. Much too often a case will be reconstructed where there will be no adjustments for the post impact braking of a vehicle in the collision, which will, by default, assume the vehicle’s braking percentage is be 100%, thereby over estimating both the post impact and impact velocities of the involved vehicles.
Based on the presence of the mark produced by the right front wheel, the fact that there was no damage to the Malibu that would
have caused that wheel’s movement to be retarded, and the mark began after the vehicles would have physically separated, it may be concluded that the brakes were being applied post collision; however, there were no other marks present from the other three wheels, nor was their any specking on the tire treads. Therefore, we may conclude the percentage of braking was something less than 100%. The problem arises as to determining how much less. If the front wheel had been locked (which it was not) from damage and there were no other marks, then it might not be unreasonable to assume that the only wheel that was braking was the right front wheel and assign a braking percentage of 35% to 45%. Nonetheless, in this case, just as if there had been no marking from any of the wheels, there is evidence that the driver applied the brakes. Therefore, simply by drawing on past experiences the braking percentage of the Malibu was estimated to be 60%...your analysis at this point may vary and that’s why this is a case study, we’re looking at a situation with known facts and evaluating our own individual analytical procedures.
Estimating the braking percentage for the Geo becomes even more difficult than estimating the percentage for the Malibu since it separated into two pieces, both of which have to be addressed as a separate entity. The easiest portion to deal with is the front of the vehicle. The post impact travel distance of the front of the Geo is so short, that virtually any braking percentage may be used while only having a minimal effect on the overall calculations. Since it is known the brakes were not being applied in the Geo after the collision, and the vehicle’s weight was no longer being supported by all four wheels, it may be conclude that post collision braking percentage of the front end of the Geo was relatively speaking “low.” The reality of the matter is, when the post impact braking percentage for the front portion of the Geo is ranged from 10% to 60% it only produced a 4.7 mph difference in the post impact velocity calculations for that portion of the vehicle. The 4.7 mph range is reduced to under a 3 mph range when dealing with the calculations for the impact velocity of the Malibu. In the calculations to follow, the post impact braking percentage for the front half of the Geo was estimated to be 25%.
Due to the fact that there would have been no braking system attached to the rear of the Geo one may surmise that the rear wheels were not locked after the collision (not to overstate the obvious). Additionally, there were no gouges or tire marks that could be attributed to the post impact travel of the rear portion of the Geo suggesting a lower post impact braking percentage. However, the rear half of the Geo traveled considerably farther than the front half which is not unexpected when looking at the damage to both vehicles. During the course of the collision, the Malibu did engage a greater portion of the rear of the Geo than it did the front. Even though the damage to both vehicles indicates that the Malibu drove through the Geo, which would have prevented the rear portion of the Geo and the Malibu from achieving a common post impact velocity, their respective post impact velocities should not have been drastically different. For the reasons cited, the post impact velocity of the rear portion of the Geo was estimated to be 50%.
So, in summary, the post impact velocity for the Malibu based on the values cited is 28 mph while the post impact velocity for the front half of the Geo was 5.2 mph and the rear half was 24.1 mph. One might immediately recognize that the “hood ornament” (the
Geo” is essentially going slower than the “hood” (the Malibu). In a normal situation, that simply can’t be; the warning message one would see in most good simulation programs would read something like “you have grossly violated ... bullet vehicle driving through target... try again...send your copy of this program back to the author, sell your computer...” But the idea here is that the bullet vehicle IS driving through the target vehicle!
Having sorted out the post-impact speeds, we should turn our attention to the relative headings and angles in this case.
Since this collision did not occur on a roadway where the configuration of the roadway and the traffic patterns might give us at least SOME clue as to the likely approach angles of the involved vehicles, an image form the overhead video was used to demonstrate the actual angles which one might compare to the analysis of the scene diagrams found on the Collision DVD. This was the only “known” data that would not have been available to an investigator at an actual crash scene that was used in this analysis and is given here as part of the solution.
Based on the videos image the approach angle of the Geo was determined to be 145deg with the approaching heading of the Malibu set to 0deg (measured clockwise following SAE J211).
Finding the departure angles of both vehicles becomes more problematic than discerning the approach angles or headings. Despite what remains a incredibly persistent total misconception, the departure angles are not measured from impact to final rest. As with the post impact speed the departure angle of the involved vehicles is defined at maximum engagement or the point of separation (of the forces). Therefore, the departure angle of the vehicles should be measured from the point of impact through the maximum engagement or separation. If a vehicle were to travel in a straight path after the collision then the departure angle would pass through the center mass of a vehicle as it sat at final rest. However, as in this case, rarely do vehicles travel in a straight line from maximum engagement to final rest.
The departure angle of the Malibu was determined to be approximately 3.2deg. When a line is drawn through the center mass of the Malibu at the point of impact through the area of maximum engagement, it still passes though the center mass of the Malibu when it is positioned along the tire scuff at a point just before it veers to the right.
If the departure angle were incorrectly drawn from the point of impact to the final resting position of the Malibu, the angle would have been misidentified by a significant amount. This error would have had a significant adverse influence on the calculated impact velocity of the Geo. By increasing the post impact heading of the Malibu, while all other values remain the same, the calculated impact speed of the Geo would increase by approximately 25 mph (!!) and the Malibu’s by approximately 10 mph.
Since the front portion of the Geo traveled for such a short distance post impact, its departure heading would not have had a chance to vary significantly from maximum engagement to final rest. Therefore, determining its departure heading did not pres-
ent a significant problem. Whereas, the departure angle of the rear portion of the Geo was not as easy to identify. As previous stated, lacking the necessary roadway evidence, establishing the area of maximum engagement becomes somewhat subjective. Based on the existing evidence the departure angle of the rear portion of the Geo was estimated to be 25-26deg.
Obviously, the departure angle of this portion of the Geo is open to debate and discussion. Nevertheless, due to the light weight of the rear portion of the Geo, the departure angle associated with the rear portion of the Geo will have the least amount of influence in the impact velocity calculations. See Figure 1.
Now, before we go more into the specifics of these angles, let’s contemplate the actual momentum analysis.
We’ve identified all of the necessary items needed to complete momentum calculations using a traditional approach: weights, post impact distances, post impact speeds, approach and departure angles. We’re left with trying to now sort out how to handle the “normal momentum analysis” where the momentum of “car two” outbound (the magnitude of speed and the direction) of the center of that mass can be approximated inasmuch as the original mass is in two parts.
There are several ways to approach this situation from adjusting the basic momentum model to solving this problem using a vector diagram. After that, we can approximate closing velocity - a particularly interesting application since this isn’t an in-line crash. The follow up to all that would be to calculate PDOF and delta-V using the facts we have available in this case.
Now, here’s the cliffhanger....at the 07 CDR User’s Conference one of the scheduled presentations is specifically on the topic of closing speed, another will be a review of delta-V data from crash tests (like this) and the content from the 07 CDR User’s conference will be the basis for the majority of content for the next issue of Collision. No, no one likes to wait that long, we’ve become a society of instant gratification, right? So, for those who have worked the problem and want answers, the actual speeds are: the Malibu was actually moving at about 36mph, the Geo at about 12 mph. Now, HOW we’d get there....that IS for the next issue!
Expert AutoStats® is a computer program containing over 36,000 vehicles cars, vans, utility vehicles, and light pickups. 1945-2006 model years are represented.
This program can be used alone or with our Expert Qwic Calcs® and Expert VIN DeCoder® programs. Also accessible from within REC-TEC.
(Laptop & Desktop IBM Compatible)
Enter: Manufacturer and Year
Select: Model from pop-up window
Output: Wheelbase Curb Weight
Overall LengthWidth
Height Drive Wheels
Wt. Distrib Front/Rear Track
Mid-60's to present output also includes (when available):
Front OverhangRear Overhang
Bumper HeightsHood Height
Turning circleWheel radius
Bumper-to-HoodBumper-to-Window
Expert AutoStats® is currently owned by over 660 private individuals and 270 Law Enforcement agencies in the U.S.A., Canada, and South America. Additional information, brochure, and order from can be downloaded from:
www.4n6xprt.com/4n6as.htm
4N6XPRT Systems®
$40.00-First vehicle, $35.00/AdditionalVehicles
$20/Additional similar Model-same Make/Year/Order
Minimum specifications include:
Weight, Length, Width, Height, Wheelbase, & Wheeltrack
VINDeCoding when VIN is providedand Information is available
Mid-60's to present also includes ( whenavailable )
Front Overhang Bumper Height
Bumper-to-hood Hood Height
Ground-to-base of Windshield
Bumper-to-Windshield
Turn
When making your request pleaseidentify:
Year, Make, Model, & Number of Doors
PICKUP/UTILITY/VAN pleasealso identify:
Short/Long wheelbase/bed2/4 Wheel Drive
Std./Super/Crew CabCargo/PassengerVan
$40.00-First vehicle , $35.00/Additional Vehicles
$20/Additional similar Model-same Make/Year/Order
You
FAX: 1-619-464-2206 or Call 1-800-266-9778 Monday-Friday 9:30 am to 5:00 pm P.S.T.
An order form can be downloaded from: www.4n6xprt.com/4n6iv.htm
PROGRAM
4N6XPRT Ped & Bike Calcs®- $ 375.00 ea.= ________
4N6XPRT StifCalcs® -$ 475.00 ea.= ________
Expert VIN DeCoder® (pkg)-$ 475.00 ea. = ________
Expert TireStuf® -$85.00 ea. = ________
California Sales Tax -7.75% = ________
Shipping & Handling = ________
S&H - Check = $11.75, Credit Card = $16.75, Govt. P.O. = $21.75
TOTAL (U.S. Dollars) = ________
Enclosed is: Visa ____ MasterCard ____ Am. Ex. ____ Exp ____/____
Card Number:____________________________________________ Name on Card:___________________________________________
Signature:_______________________________________________
Please Make Checks*/Purchase Orders Payable to:
4N6XPRT Systems®
8387 University Ave - La Mesa, CA 91941-3842 - U.S.A. FAX: (619) 464-2206 Phone: (619) 464-3478
Web: http://www.4n6xprt.com
Checks *MUST* be drawn on a U.S. bank. Orders are shipped within 10 working days of receipt. Prices subject to change without notice. Multiple orders and Package discounts available. Please call for details.
10 Collision: The International Compendium for Crash Research
(c)2006CrashDataGroupInc
puts the NHTSA Crash Test database on YOUR 32 bit MS-Windows computer PLUS calculates A-B-G values and Test Specific Crush Factors based on the NHTSA test data, with no need for you to get on the internet, all for the price of only $475.00*!
For more information and sample printouts please visit our web site at http://www.4n6xprt.com You may also call us at (619) 464-3478 M-F, 9a-5p (PST), if you have specific questions.
4N6XPRT Systems
http://www.4n6xprt.com
8387 University Avenue - La Mesa, CA 91941 - USA
Phone: 800-266-9778 / (619) 464-3478 Fax: (619) 464-2206
Individual Vehicle Stiffness Data
You want Stiffness data for a vehicle, QUICKLY & CONVENIENTLY, but you don’t want to buy a program? Order your data on a vehicle-by-vehicle basis from us at only $40.00 per test!
Year: _________ Make:___________________________
Model:_________________________________________
Impact on (circle one): Front - Side - Rear
Impact speed (circle one): High - Low YOUR CONTACT INFORMATION
Name:_________________________________________
Company_:______________________________________
Address:________________________________________
City/State/Zip_:___________________________________
Phone_:_________________________________________
Fax::__________________________________________
Visa/MC/AmEx_:_________________________________
Exp.:____/____ Case Reference_:____________________
Send Order to: 4N6XPRT Systems
Fax: 619-464-2206 Ph: 800-266-9778
8387 University Avenue / La Mesa, CA 91941 / USA
(c)2006CrashDataGroupInc
November ‘06
11/5/2006 - 11/9/2006 Jacksonville, FL
CRASH ZONE
IPTM
12/4/2006 - 12/8/2006 Panama City, FL
PEDESTRIAN/BICYCLE CRASH INVESTIGATION
IPTM
11/6/2006 - 11/10/2006 Evanston, IL
Traffic Accident Reconstruction 2 Northwestern University
11/13/2006 - 11/15/2006 Evanston, IL
Pedestrian Vehicle Traffic Collisions Northwestern University
11/13/2006 - 11/17/2006 Tempe, AZ
HUMAN FACTORS IN TRAFFIC CRASH RECONSTRUCTION
IPTM
11/12/2006 - 11/16/2006 JACKSONVILLE, FL
TRAFFIC CRASH RECONSTRUCTION UPDATE
IPTM
11/16/06 - 11/18/06 S. Lake Tahoe, CA
Crash Reconstruction, Start to Finish
CAARS
11/27/2006 - 11/29/2006 Pittsburg, CA
Crash Zone - Basic
WeCARE
11/27/2006 - 12/1/2006 Jacksonville, FL
COMPUTERIZED COLLISION DIAGRAMMING
IPTM
11/30/2006 - 12/1/2006 Pittsburg, CA
Crash Zone - Advanced
WeCARE
12/4/2006 - 12/15/2006 Key West, FL
AT-SCENE TRAFFIC CRASH/TRAFFIC HOMICIDE INVESTIGATION
IPTM
12/4/2006 - 12/15/2006 Albany, NY
TRAFFIC CRASH RECONSTRUCTION
IPTM
12/4/2006 - 12/15/2006 Concord, CA
Advanced Accident Investigation
WeCARE
12/11/2006 - 12/15/2006 Tempe, AZ
TRAFFIC CRASH RECONSTRUCTION UPDATE
IPTM
12/12/2006 - 12/12/2006 Atlanta, GA
CDR Technician Course Collision Safety Institute
12/13/2006 - 12/15/2006 Atlanta, GA
CDR Data Analyst Certification Course Collision Safety Institute
1/8/2007 - 1/12/2007 Evanston, IL
Vehicle Dynamics
2/1/2007 - 2/2/2007 Houston, TX
Northwestern University
1/8/2007 - 1/12/2007 Johnson, IA
Vehicle Dynamics
Northwestern University
1/8/2007 - 1/12/2007 Jacksonville, FL
APPLIED PHYSICS FOR THE TRAFFIC CRASH INVESTIGATOR
IPTM
1/8/2007 - 1/12/2007 Pittsburg, CA
Computer Assisted Diagramming
WeCARE
1/15/2007 - 1/26/2007 Evanston, IL
Traffic Accident Reconstruction 1
Northwestern University
1/16/2007 - 1/16/2007
Largo, FL
CDR Technician Course
Collision Safety Institute
1/17/2007 - 1/19/2007
Largo, FL
CDR Data Analyst Certification Course
Collision Safety Institute
1/22/2007 - 1/26/2007
Panama City, FL
INVESTIGATION OF MOTORCYCLE CRASHES
IPTM
1/29/2007 - 1/31/2007
Houston, TX
2007 Crash Data Retrieval User’s Conference
ARC Network, Collision Safety Institute, Vetronix,
1/29/2007 - 2/2/2007 Evanston, IL
Traffic Accident Reconstruction 2
Northwestern University
CDR Train the Trainer Course (Technician Course) Collision Safety Institute
2/5/2007 - 2/9/2007 Jacksonville, FL
PEDESTRIAN/BICYCLE CRASH INVESTIGATION
IPTM
2/5/2007 - 2/16/2007 Concord, CA
AT-SCENE TRAFFIC CRASH/TRAFFIC HOMICIDE INVESTIGATION
IPTM
2/5/2007 - 2/16/2007 Miami, FL
ADVANCED TRAFFIC CRASH INVESTIGATION
IPTM
2/12/2007 - 2/16/2007 Jacksonville, FL INVESTIGATION OF MOTORCYCLE CRASHES
IPTM
2/19/2007 - 3/2/2007 Jacksonville, FL TRAFFIC CRASH RECONSTRUCTION
IPTM
2/20/2007 Stillwater, OK CDR Technician Course Collision Safety Institute
2/21/2007 - 2/23/2007 Jacksonville, FL CDR Analyst Course Collision Safety Institute
3/5/2007 - 3/9/2007 Jacksonville, FL
4/2/2007 - 4/6/2007 Evanston, IL
HUMAN FACTORS IN TRAFFIC CRASH RECONSTRUCTION
IPTM
3/5/2007 - 3/9/2007 Jacksonville, FL
CRASH ZONE
IPTM
3/5/2007 - 3/16/2007 Evanston, IL
Accident Investigation 1
Northwestern University
3/12/2007 - 3/23/2007 Miami, FL
ADVANCED TRAFFIC CRASH INVESTIGATION
IPTM
3/12/2007 - 3/23/2007 Concord, CA
Traffic Accident Reconstruction I WeCARE
3/19/2007 - 3/30/2007 Johnson, IA
Traffic Accident Reconstruction 1
Northwestern University
3/19/2007 - 3/30/2007 Evanston, IL
Accident Investigation 2
Northwestern University
3/19/2007 - 3/23/2007 Abington, PA
PEDESTRIAN/BICYCLE CRASH INVESTIGATION
IPTM
Basic Physics and Mathematics Workshop
Northwestern University
4/2/2007 - 4/6/2007 Jacksonville, FL
TRAFFIC CRASH RECONSTRUCTION UPDATE
IPTM
4/9/2007 - 4/20/2007 Jacksonville, FL
AT-SCENE TRAFFIC CRASH/TRAFFIC HOMICIDE INVESTIGATION
IPTM
4/16/2007 - 4/20/2007 Miami, FL
PEDESTRIAN/BICYCLE CRASH INVESTIGATION
IPTM
4/16/2007 - 4/27/2007 Evanston, IL
Traffic Accident Reconstruction 1
Northwestern University
4/23/2007 - 4/27/2007 Jacksonville, FL
SPECIAL PROBLEMS IN TRAFFIC CRASH RECONSTRUCTION
IPTM
4/23/2007 - 5/4/2007 Jacksonville, FL
ADVANCED TRAFFIC CRASH INVESTIGATION
IPTM
4/30/2007 - 5/4/2007 Evanston, IL
Traffic Accident Reconstruction 2
Northwestern University
5/7/2007 - 5/9/2007 Pittsburg, CA
Crash Zone - Basic
WeCARE
6/18/2007 - 6/22/2007 Jacksonville, FL
PEDESTRIAN/BICYCLE CRASH INVESTIGATION
IPTM
5/7/2007 - 5/11/2007 Evanston, IL
Heavy Vehicle Crash Reconstruction
Northwestern University
5/7/2007 - 5/18/2007 Jacksonville, FL
TRAFFIC CRASH RECONSTRUCTION
IPTM
5/10/2007 - 5/11/2007 Pittsburg, CA
Crash Zone - Advanced
WeCARE
5/14/2007 - 5/18/2007 Evanston, IL
Photogrammetry for the Accident Reconstructionist
Northwestern University
5/14/2007 - 5/18/2007 Jacksonville, FL
ADVANCED TRAFFIC CRASH RECONSTRUCTION WITH WINCRASH
IPTM
5/14/2007 - 5/18/2007 Concord, CA
Intermediate Accident Investigation
WeCARE
5/21/2007 - 5/25/2007 Johnson, IA
Traffic Accident Reconstruction 2
Northwestern University
5/21/2007 - 5/25/2007 Miami, FL
INVESTIGATION OF MOTORCYCLE CRASHES
IPTM
6/18/2007 - 6/22/2007 Jacksonville, FL
ENERGY METHODS AND DAMAGE ANALYSIS IN Crash RECONSTRUCTION
IPTM
6/18/2007 - 6/29/2007 Concord, CA
Advanced Accident Investigation
WeCARE
6/19/2007 - 6/21/2007 Evanston, IL
Traffic Accident Reconstruction Refresher
Northwestern University
6/25/2007 - 6/29/2007 Jacksonville, FL INVESTIGATION OF MOTORCYCLE CRASHES
IPTM
6/4/2007 - 6/7/2007 Las Vegas, NV
ARC - CSI Crash Conference
Accident Reconstruction Network, Collision Safety Institute
(c)2006CrashDataGroupInc
If you would like to list an event in future issues, please send an email with complete information to editor@collisionpublishing.com
(c)2006CrashDataGroupInc