Project SC300 Road Racer: Part 31 – ECM(easure all the things!)
bottom of hubstand on scale pad and leveler and tape marks on the concrete below
Here’s the rear centerline transferred over 4-feet.

You’ll also notice that the hub center was transferred onto the floor. I was able to use the Intercomp laser level system to do this. The laser features both a horizontal and a vertical laser mode. With the laser in the vertical mode, I ensured that the laser beam hit the center of the wheel hub. I then moved the laser body to ensure that the beam was aligned with a square (the tool) coming off the vehicle center offset line.

The line that is perpendicular to the vehicle center offset line that intersects the center of the hub is the center of the “axle.” The midpoint between both “axle” lines is the center of the wheelbase (half wheelbase). The intersection of the 4-foot offset line and the half wheelbase is the exact center of the suspension, offset by 4-feet.

 

mans hands aligning a level laying on the concrete with a long piece of tape, square sitting to the left
Here you see a mark for the half wheelbase point as well as the four-foot line.

Another thing to note is that, if you’re concerned with hyper accuracy, even the permanent marker’s width potentially introduces an offset. If you’re capable of measuring 0.0001″, that 2mm marker line is HUMONGOUS. Here we are using the very edge of the masking tape to represent the line and not a drawn line. We did end up making several dots on this line for repeatability, but we used the edge of the tape itself for touching with the probe.

Given all the other slop here, is this extra step worth it? Maybe, maybe not.

In sims as in real life, the tire is the only thing touching the road. Generally speaking, all consumer-grade sim racing tire models suck. Some are better than others. To get something as good as rFactor can provide, I would have to purchase another Nankang AR-1 and dissect it to take internal measurements, figure out spring rates of the sidewall, and many other factors. This would cost a tremendous amount of time and money. At my level, Nankang would never share these details with me, so dissection and measurement on my own would be my only option.

You may not realize it, but an assumption is also baked into all of this about the track surface modeling in the simulator. If the track was made from a 3D laser scan (using ECM equipment??), the surface detail would be extremely accurate. Most of the tracks I drive here in the South East were not produced by Studio 397 (the developer of rFactor 2) and are not laser scanned, which means all bets are off as to how faithfully the track model reproduces the available real-world traction.

Given the tire inaccuracy and the track surface inaccuracy, If I can end up with a model for the sim that reproduces my real-world times within several seconds and has even remote resemblance to the driveability and dynamics of the real car, I will call it a win.

Is the effect on the repeatable accuracy of a permanent marker’s width going to be the difference between 3 seconds off my real-world pace and 15? Probably not. But it eventually adds up. So, do what you can within reason.

Onwards.

 

large, flat, round foot of tripod sitting next to a piece of tape with a clearly traced mark
Remember, we talked about the base being the reference for the portable CMM itself?

We marked each of the feet of the base as best we could, and tried to make sure that the tripod stayed put. The base isn’t all that heavy, and if you are ham-fisted with the arm, you will inevitably move it. If you’re just generally a ham-fisted klutz, you will probably bump into it and move it.

If you want additional accuracy, predictability, and stability, use heavy sandbags or, better still, bolt it to the floor. There are also other tricks you can employ to ensure consistent calibration throughout your measurements, like gluing reference points all over the place and continually returning to them to re-calibrate. ECM can help you here, too.

Now, finally, for the measuring!

If you’re terrified at using a computer, let alone using CAD-type software to take precise measurements, don’t fret. ECM can offer training and remotely help throughout the process. You could probably even hire them to come and do it for you, but that starts getting expensive and eating into any potential return on your investment (if you’re going down a manufacturing route). If you’re planning to do this often, the investment in training or assistance could pay off in the long run (versus the time and frustration of learning the hard way).

As an example, ECM was even able to help me on a Saturday afternoon at 4PM when I was trying to figure out how to use the software. It’s all part of the rental.

But maybe don’t wait until 4PM on Saturday to try to use it. I’m sure they’d prefer your call during regular business hours. Being “on call” stinks, even when it is your job.

OK, really, it’s time for measuring!

11 comments

  1. There’s places that will do tire testing and it’s… expensive but if you keep the test matrix down, in the 4-figures level. Would be interesting to add that to the mix. I’m figuring that most of the time in sims it’s not that the tire model itself has issues, just that it’s filled with variables that people are guessing at.

    1. The most important thing is that the driver understands how the tire develops grip, and how to keep it at the proper temp. You can collect as much data as you want, but if the driver doesn’t understand that data, it is worthless. The pneumatic tire has been around for over a hundred years and there’s only one book written about it? Sad.

      This is a must read for anybody serious about racing:

      https://www.sae.org/publications/books/content/r-351/

  2. Very interesting, great work Erik! I had no idea this type of equipment could be rented and used by near mortals.

    The wheels are turning on other ways this type of equipment could be used.

    1. Calspan does – if you are willing to be flexible on schedule and don’t get too complicated it helps hold the cost down, and the gent I was talking to was willing to deal with it as a smaller project.

  3. Yeah, I really like this article. I wrote my own code in MATLAB in college using Dixon’s equations and precise measurements (by hand) of the EG6 Civic suspension. I had so many parts lying around, it was pretty much a no brainer. You can learn how to avoid so many mistakes through simulation. While I did some minor modifications using a LCON traction bar which allowed me to remove the rear half of the LCA, and drop 8lbs of unsprung weight, I also gained enormous appreciation for Honda suspension engineers. The stock bump steer curve could not actually be improved upon without new spindles. Simulation proved to me that some things are better left alone.

    This is a great book, BTW. It is the ‘Bible’ for suspension engineers.

    https://www.wiley.com/en-us/Suspension+Geometry+and+Computation-p-9780470510216

    1. Thanks, Joe! That looks like a real textbook there. Probably well over my head. As you’ve learned some things are best left alone, I have learned that for some things I should probably outsource. Suspension engineering is one of those things. I’m fortunate to have folks like Figs Engineering, Mike Kojima, and Rob Lindsey (Morlind Engineering) available to me!

Leave a Reply

Your email address will not be published. Required fields are marked *

*
*