Turning Right better than Left (DA?s)

[Chris_Swearingen]

Mystery Solved. Bad instrumentation. At the small test and tune after the July 22 event, I borrowed (thanks Rob) another gtech and used it for measurments. It came out to within 0.01

I had convinced myself that there was something wrong with the car and I had "data" to back it up.


Moral of the story: Be sure to validate the data before acting on it.




Can anyone who collected data Saturday and Sunday share some numbers to help me confirm that my Miata does indeed turn to the right better than the left?

I show an average of .199 more Gs turning right than left for both days.
Saturday was .180 more and Sunday .219 more.

I want to compare with someone else to make sure it wasn't just a course difference, but I show about that same difference over all the runs/events starting 5/27.

It was really noticeable to me on Saturday to the extent that I thought I had an alignment problem. I checked the toe, and verified the alignment wasn't the issue (thanks for the help Paul). I "compensated" the rest of the weekend by being careful in left hand turns, but data shows a trend.

If the data proves out, the next question will be, how do I remedy this and get equal turning ability both left and right?

Any thoughts would be appreciated.

[MrPickles]

It sounds to me, that possibly you might have an issue with front shocks. When was the last time they were re-valved or Dyno'd?

[Chris_Swearingen]

It sounds to me, that possibly you might have an issue with front shocks. When was the last time they were re-valved or Dyno'd?

Long, long ago, in a galaxy far, far away....

Good thought. I will see about checking those. Anyone else have a thought?

[S Abshire]

I would take those nubers with a grain of salt, all of the corners are different so right to left gs would also be different. You need to get on a skid pad and run it one way and then the other if you want to compare left vs right turning.

[Chris_Swearingen]

I would take those nubers with a grain of salt, all of the corners are different so right to left gs would also be different. You need to get on a skid pad and run it one way and then the other if you want to compare left vs right turning.

Scott,
I agree that one course etc is not a good indication, but I see that trend across several events/courses. I was looking for a comparison of someone else's max gs for Saturday and Sunday to see if anyone else had the same spread. If so, I would be less concerned, but I would guess that is not the case. I am not asking for actual numbers, I don't care what anyones max g value was, I only care if they had a delta between left and right on both Saturday and Sunday and if so, what that delta was.
--Chris

[S Abshire]

I'll look at my data over lunch and see what I come up with.

[Chris_Swearingen]

Thanks!

[S Abshire]

Saturday I showed a .15g bias with left hand turns being higher. Sunday I showed right hand turns being higher by .04g's.

Now my unit is doing some sort of fuzzy math to figure out lateral g's, because it is all GPS based as the max numbers seem relatively low to me, but then again I dont remember the car laying over on one side or the other as it seemed to just slip through the corners.

I would wonder if it has anything to do with placement of your da unit? Is it dead center? level?

[Chris_Swearingen]

Pretty much dead center and level, although I don't recall that being an absolute requirement in the installation instructions. Your values seem to confirm to me that I have an issue. Thanks for sharing.

[Mark C]

Its just a thought but have you checked your corner/cross weights? You may have more on one diagonal than the other. As I recall, the cross weights being equal should help the car corner equally in both directions where a weight jacked car with uneven diagonals will bias one way or the other.

-Mark

[Chris_Swearingen]

Corner cross weights are close enough that I don't think that is the issue.

http://www.systemstrategies.com/miata/CornerWeights.JPG

[PaulL]

When I used to log data with the Geez Cube more often, I never saw any significant difference between max right/left g's ... usually there was at least one turn of each sort on the course that you could get a good sustained g number, and left/right were always very close.

Did you shake and try to move the suspension? Although the static alignment was fine when we checked it, maybe you're getting some dynamic change.

Did you get the thrust alignment done, too? That was something we didn't check ... maybe it's pointing off center.

Could it be diff related, I wonder? ....

And are your tires in similar condition all around?

[LoCore]

At PMI last month, I had the distinct impression that the car was turning right much better than left.

Data acquisition confirmed it.

Eventually found that the rear compresssion was set different on each side (had half of a turn more rebound in the rear right).

Coincidence? Don't know. But the suspension settings will be equal for this weekend :-)

[Chris_Swearingen]

Nothing dynamic appears to be going on, and when I review the data it seems to have been going on for a while. The new tires and alignment settings just make it more obvious. The spread isn't any bigger I can just feel it better.

Thrust angle is 0.01 degrees. Rear toe measured -0.01 and -0.03 left and right for just the tinyest bit of toe out. I don't think I am a good enough driver to notice -0.02 degrees more toe.

Shocks have been mentioned as a possible culprit a couple of times. Anyone have a set of freshly revalved double adjustables for a Miata they want to get rid of cheap?:)

[khammack]

Something that I don't think anyone has mentions is driver preference. Is it possible, that you are just more conformable turning one way better than the other?

[Chris_Swearingen]

I don't think so. On my third run Saturday when it was most pronounced to me, it felt really good turning right and would slide turning left. There are many many issues that can be correctly blamed on the driver, but I don't think this one was caused by the loose nut behind the wheel.

[aleliaert]

Some theories...

Perhaps surface type and/or camber played a role? World Arena has that one area near the paddock with unsealed asphalt where on-camber sweepers tend to be.

And are you comparing momentary peak Gs or sustained numbers through sweepers? From our DL1 data the peak left/right numbers can be off a few percent, but a look at the graphs suggests that the "visible average" is roughly the same.

One other idea - are you sure you're staying off the bumpstops? A sudden spike in your spring rates as the car is settling into a corner can cause some handling problems. With those fat new Hoosiers, maybe cornering forces have increased and you're in need of more suspension travel and/or more spring rate. Are you running special mounts and/or shortened Konis?

[Chris_Swearingen]

We ran the section you are referring to both directions. Both peak and steady state numbers show the difference.

I switched the front shock mounts out for NB mounts to gain some travel. I think the backs had enough travel. I will pull the wheels off at the next event and take a look at where the zip ties on the shock shafts wound up. Another good thought.

Spring rates are at 650 and 400 to compensate for the new tires.

Paul you and Evan both drove it Saturday, did you feel it get into the bumpstops? I never did.

Thank you all for your input, keep the ideas coming.

[PaulL]

Nope, never did hit the bumpstops.
Did you record data during my run?

[Chris_Swearingen]

Nope, never did hit the bumpstops.
Did you record data during my run?

I wish I had, but no, I didn't record either of you. (Note to self, record all runs no matter who is driving)

[S Abshire]

I wish I had, but no, I didn't record either of you. (Note to self, record all runs no matter who is driving)

If you let Paul or Evan drive also remember to "turn it on". Ask me how I know. I didnt even go 2 weeks without blowing rule number one of the DA school.

[Chris_Swearingen]

Mystery Solved. Bad instrumentation. At the small test and tune after the July 22 event, I borrowed (thanks Rob) another gtech and used it for measurments. It came out to within 0.01

I had convinced myself that there was something wrong with the car and I had "data" to back it up.


Moral of the story: Be sure to validate the data before acting on it.