Tyre defelction modelling

[Karen]

I am from Queen's University Belfast, my final year project is based on modelling tyre deflection and tyre grip. Does anyone know of any packages that would be useful in producing a 3D model of this? I have treid Cosmos Motion but it is not proving very successful!

[Karen]

I am from Queen's University Belfast, my final year project is based on modelling tyre deflection and tyre grip. Does anyone know of any packages that would be useful in producing a 3D model of this? I have treid Cosmos Motion but it is not proving very successful!

[John_Burford]

You should be more specific about what you are trying to accomplish. Modeling the behavior of a tire is more of a career, not a project.



John Burford

[Schumi_Jr]

I'm also not sure waht you're trying to do, but this link might be of use to you - http://www.abaqus.com/solutions/tb_pdf/TB-03-TRT-1.pdf

It's a bit of a sales pitch but still interesting...

[Dave M]

for $500 bucks you could buy some tire data from the tire consortium, seach for it.

[jdstuff]

Karen,

I apologize in advance if I sound a bit cynical, but I hope that you realize the complexity of what you are requesting. Having worked for a tire company for a few years, and even spending time in the modeling/simulation group, I can tell you there isn't any "off the shelf" software that will accomplish what you are asking.

Generally speaking, much of the modeling software utilized by major tire companies is developed in house. Sure, they often use a number-crunching powerhouse like Abaqus to solve the mathematical model....but all of the important stuff like the pre/post processing, mesh generation, etc. is done by software that they have written themselves.

One thing that you must realize is that a typical passenger car tire can contain around 25 different materials. Many of these are different rubbers, who's material properties are non-linear. The rubber components are then reinforced with a matrix of fibers that is comprised of steel, polyester, nylon, aramid, and sometimes even fiberglass. All of which whoes material properties are drastically different than rubber. Finally, the entire structure is supported by air....which again acts as a non-linear spring. This still does not even begin to take into account the thermal properties of all of the materials (nylon will shrink when heated while rubber will expand, etc.), or how all of these non-linear materials interact at their respective boundaries.

At the particular company which I worked for, the only predictive tire models that were considered reasonably accurate were those with static loading conditions. For example, placing a normal load on the tire in order to predict deflection and footprint shape. Some of these models were complex enough to tie up an entire weekends worth of processor time on the company servers! As far as dynamic models go, and predicting things such as tire grip, they had only begun to scratch the surface...nothing had yet been developed that would be considered accurate enough for production use. Goodyear has a pretty advanced modeling group, and they have worked directly with Sandia National Laboratories (http://www.sandia.gov/) .....they are a government agency that have perhaps some of the post powerful supercomputers in the United States. Some of the 3-D dynamic models that they created ran for WEEKS on Sandia's mainframes!

Anyway, I think your best bet might be to consult with a tire company and seek their assistance....if that is possible. Best of luck with your project!

[ben]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Schumi_Jr:
I'm also not sure waht you're trying to do, but this link might be of use to you - http://www.abaqus.com/solutions/tb_pdf/TB-03-TRT-1.pdf

It's a bit of a sales pitch but still interesting... </div></BLOCKQUOTE>

I hate to join the people suggesting this is a blind alley for a single-year Uni project, but to be honest, it is.

The ABAQUS link is good and does show what is perfectly possible, but as Jason has pointed out, adequately positioning all the reinforcing bar elements is not something standard ABAQUS will allow you to do easily. We have proprietry in-house pre-processing to do it. Material properties are another big problem which Jason alluded to. We have problems even compared to road car modelling because of different temps and the difficulties in actually testing compound properties of very soft compounds.

Ben

[Kevin Hayward]

Karen,

I think it would be in your best interests to thoroughly check what your project is supposed to be about. If it is supposed to be something similar to FEA modelling of a tyre then you might as well forget it ... as mentioned by Ben and Jason.

However if your project is about estimating tire deflection around a track and/or testing it then there are a few things you can do to get a reasonable answer. These include dynamic car models using existing tyre data (maybe from the consortium) as well as testing with ride height sensors and shock potentiometers.

This sort of knowledge can help you assess the effects of tyre deflection on the roll moment distribution and estimates of the load variation on the tyre itself. With the testing information you can use it to estimate the actual rolling stiffness of the tyre, changes to stiffness due to camber (with kinematics data) as well as correlations between tyre deflection and tyre temps (with temp sensors).

I would bet that there is a lot you can learn with these sensors. However I have not tested with ride height sensors, so haven't had the opportunity to make the most out of it. I am pretty sure that Monash did some testing like that a few years ago and the guy was working on some pretty amazing tyre models. He was also a postgraduate student.

Kev

OptimumG