Z-man

[quote]Originally posted by Matt Olde:
<strong>&lt;snip&gt; Fork over the $125, slap that puppy on, take your baby for a drive, and know you're probably getting the most bang-for-performance out of any benjamin you've spent on your p-car ever.

Merry Christmas!
-Matt</strong><hr></blockquote>
Spend another $25 ($150 total) for an M030 968 19mm rear sway bar: that, IMHO, will yield a better improvement in handling for our cars. (I know, spoken like a tru nay-sayer! )

Anyway: hplj123: I have a question for you: If I'm reading this correctly, your test was done on an attached brace. I would like to see what the results would be if there was no bar (or if the bar was loosened) to see how much more lateral movement would occur. Don't know if your test can measure that, but unless you can show a comparison between non-STB and STB, then you haven't proven anything. What if without a STB, there's only 8-10mm of flex? That's probably not a noticable amount of difference!

Good job on the test: I just want to see a comparison. Keep digging!

-Z.

slevy951

[quote]Originally posted by Z-man:
<strong>
Anyway: hplj123: I have a question for you: If I'm reading this correctly, your test was done on an attached brace. I would like to see what the results would be if there was no bar (or if the bar was loosened) to see how much more lateral movement would occur.
-Z.</strong><hr></blockquote>

Answering for Ken because I was present for the test. The bar was loose in the pipe. It could slide laterally within the pipe. Basically we were trying to see what movement there was without the brace. I think we've proved there is movement.

Z-man

[quote]Originally posted by slevy951:
<strong>Answering for Ken because I was present for the test. The bar was loose in the pipe. It could slide laterally within the pipe. Basically we were trying to see what movement there was without the brace. I think we've proved there is movement.</strong><hr></blockquote>
Ok: NOW see how much movement there is when the bar is tightened down. I guess you'll have to fabricate something that is loosely connected to both strut towers next to the strut tower bar and be able to measure the flex.

BTW: 6mm is a really small amount of movement, if you ask me. However: 6mm of movement on top may translate into more movement down below. (Next test: develop something that measures flex down low at the a-arm level: with and without a STB!)

Keep on going: you guys are doing a great job! (But you still haven't convinced me to get a STB! hehe...)
-Z.

slevy951

[quote]Originally posted by Z-man:
<strong>
Ok: NOW see how much movement there is when the bar is tightened down. I guess you'll have to fabricate something that is loosely connected to both strut towers next to the strut tower bar and be able to measure the flex.

BTW: 6mm is a really small amount of movement, if you ask me. However: 6mm of movement on top may translate into more movement down below. (Next test: develop something that measures flex down low at the a-arm level: with and without a STB!)

Keep on going: you guys are doing a great job! (But you still haven't convinced me to get a STB! hehe...)
-Z.</strong><hr></blockquote>

Yeah, with the bar attached is harder. We're thinking we'd have to attach the marker/measuring device to the center of the firewall. Not sure how that's going to get accomplished.

6mm is alot. If you put one of ours on your car, you will feel the difference. I was just as skeptical as you until I got one on my car. You can buy one and try it. I'll give you 30 days with it. If you don't like it, send it back in the original box and in resellable condition. I'll refund full amount. I'll even cover the return shipping for you.

SidViscous

Putting onthe firewall is a false test. Because now you have additive movement. Movement of the strutbrace and movement of the firewall.

That's the whole problem with trying to do this test with $6 of parts from Home Depot. You have to eliminate other factors so that all you see is what you want to look at. Everything else is noise. If your data barrier doesn't rise above the noise you have crap data.

hplj123

You guys are wearing me out. I"m going to try and answer several people's questions with one post.

Michael.

I did look at the link to the Camber Truss that you provided. It looks like a great bar. It appears that it will cost over $400.00 to put this bar on a turbo or S2.
Part No. Model Price

CT 944 944, 944S $239.00
CT 944T 944 Turbo, Turbo S, S2, Top Bar Only $159.00


The KLA total bar shipped is still less than the Top Bar Only price and this bar has much greater weight. I hate trying to discuss someone else's product.
Scott,

I'll make you a set of solid ends if you want but I think the improvement will be minimal. I also don't want to take any responsibility if this causes cracks somewhere else in the chassis. I have not taken the time to do a Failure Mode and Effects Analysis (FMEA) on the stresses. As I said, the ends will cost about $30.00 each and will take 3 to 4 weeks to complete. I would doubt that they would make much difference unless you are running slicks in a very aggressive situation.

M758

Hi. I ran ITS for many years driving a 280Z. Keep us all informed as to how you fair. I would like to adders the issue you brought up about the simple sharpie test I ran. I accounted for pin movement when I said the results were +- 1mm. I can assure you that the end of the pin was fit snugly into the pipe so it had no lateral movement. I also taped the pin so it would not move from side to side. If you look at the results you can see where the mark was in on either side of the center point. The big spot is the center and the mark goes about the same distance on either side. I cannot imaging that we would get over ¼” of vibration movement and I can assure you the pin was in tight and did not move. I think this test shows that the towers do flex.

Sid.

$50,000 worth of equipment to check the effect of a $125.00 bar may be a little bit of overkill although I would be interested in the final result.

Z Man

I did the test with the bar totally disconnected on one side and use the pipe as a slip joint. The 6mm movement was the spread on the towers off their true vertical position. The way the suspension is set up it will make a difference in the camber of the tire. If camber is changing on one tire more than the other you will have a camber offset or spread. If you have a camber plate try moving the camber of one tire 4 or 5 degrees and leave the other alone. Now try Driving the car and see if it changed the handling. That is what you are facing without a bar. Most daily drivers will never need the added strength. But, for those of us who drive our cars hard it will make a difference in feel. I was surprised over the difference it made on my S2 on the street. It also made a bid difference on the Turbo.

I agree that the slip joint was not science. I am in Engineering and Quality is an aerospace company that designs and builds advanced composites so I do understand the theory behind how to test. To get an undisputed result would require the kind of instrumentation that Sid is talking about. In the end it will show that the flex between the towers is greatly reduced. Then all we will have left to debate is does the reduction in flex have an effect on the drivability of the car.
Finally to everyone. I have taken a look at the strut tower as it is attached to the body. The concern of forward and aft movement makes little sense to me because of the attachment method. First the strut tower is attached to the longitudinal. To get forward and aft movement means the longitudinal would have to flex up and down. The boxed construction would prevent that. The flex we are concerned about is in and out which will apply a twisting moment where the boxed construction is weakest. Forward and aft movement is trying to bent the boxed longitudinal in its strongest direction. In addition to that the strut and the inner fender are welded in such a way as to provide additional triangulation and strength. This simply won’t happen. This is the same issue for vertical movement. For the tower to move vertically it would have to bend the longitudinal in its strongest direction. That won’t happen.

The weakest direction of movement that we are trying to eliminate (or at least reduce) is flex in and out. The bar is not designed to make a rigged box. It again is simply allowing both towers to hold all of the force instead of one side carrying more. For those of you who have raced a car and put many hours one can attest, rigged it great but is does result in cracks developing in welds and around holes. A little flex in not totally a bad thing but a lot of flex is.


I hope this answers several questions. I’m not sure I have the energy to keep up with this debate, but I’ll try.

Ken

SidViscous

$50,000 worth of equipment to check the effect of a $125.00 bar may be a little bit of overkill although I would be interested in the final result."

No-one said anything about paying that much money. Thats what I can borrow for free. Equipment we need to buy can be kept under $100.

And it's not just the effects of the $125 bar. It's how the car reacts to high lateral loads in real world conditions we want to test. The bar is simply how we would solve it, if it were a problem.

slevy951

[quote]Originally posted by SidViscous:
<strong>Putting onthe firewall is a false test. Because now you have additive movement. Movement of the strutbrace and movement of the firewall.

That's the whole problem with trying to do this test with $6 of parts from Home Depot. You have to eliminate other factors so that all you see is what you want to look at. Everything else is noise. If your data barrier doesn't rise above the noise you have crap data.</strong><hr></blockquote>

I agree. I don't know how to do it otherwise within a reasonable cost.

I think we agree there is movement between the towers and we agree that an STB helps. I think the question now is what is the difference between a rigid fixed joint and a heim joint in this application, if any.....

Doug Donsbach

Since the only real measure of a performance part is how much performance you get from it, I don't see a lot of useful data coming from your test.

Now, if you tell me that you plan to go to a track and turn times back-to-back, assuming you can turn repeatable laps and control the other conditions (tire pressures, ambient and track temperatures, fuel load, etc.) - that would be useful information.

SidViscous

"Now, if you tell me that you plan to go to a track and turn times back-to-back, assuming you can turn repeatable laps and control the other conditions (tire pressures, ambient and track temperatures, fuel load, etc.) - that would be useful information."

You can't. That's the whole problem. That's why you look for the kind of data that will tell you something about what the car is doing, make some educated geuses, and hope that the combination of things will give you better lap times. You wan't better lap times alone, then put in a more powerful motor, barring that you try for little tweaks.


To give you an example. My experience with strain gauges comes largely from working with Dams. They measure stress in the dam, try to figure out where a cut will relieve the most stress, while keeping the integrity of the dam. Then they bring in a saw and cut it at a cost of about 5 million dollars.

Under your methodolgy you would take a geuss, cut the dam, see what that does, then cut it again somewhere else and see if it gives you a "Better lap time"

If you really want to fine tune the cars setup you have to understand what is going on, then see what you can do to minimize issues within your budget.

hplj123

Sid,

Please don't take my reply as a slam, I would love to see the results of any test and was only saying to get a true result would take about 50K worth of insturmentation. if you can get that done i think we would all love to see the results. I know I would. If you can do that the YOUR THE MAN. please let me know how I can help.

Ken

M758

Sid,
I completly understand your Dam analogy to a point. It works for Dams, but not for cars.

Quite simply the best method is to do back to back tests with and experinced driver. This will measure the total performance of the bar by measuring the resultant change in car performance. The reason I stick by this method is that even if you know exactly how much less movement the bar allows how do you translate that in to the bottomline whichs is car performance? To me even the most talented engineers do not have all the equations to do calcuations like this and lap time tests are very important. Even the guys at the Ferrari F1 do alot of on track testing. Why? Because when the day is done that is all that really maters. Rememeber that while the brace does stiffen the chassis it also addes weight in bad place. High and at the front. Does the effect of the extra weight counter balance the extra stiffness? Not sure. Probably can only be sure by testing since the equations are so complicated.

Guys..
Don't take this too mean your test was stupid and worthless. It is not, but it does not provide a detailed answer to the question about performance. True you can test the KLA bar vs all others and with a well thoughtout strain gage test can show that the KLA bar is xx% stiffer than the rest and produceds less tower movement. But again we come to the qestion of "So how does that make my car faster?"

SidViscous

I didn't mean it as a terse reply and I didn't think it was a slam. I just wanted to point out that we didn't have to raise 50K to do the test. The people I consult for have said "Anything we have that you need your welcome to."

Unfortunately I haven't even got the $100 right now, nor do I have the facilities to make a test bar, nor a car that would be good for testing (need motor mounts bad) or a closed course to do the testing. I can bring the other equipment to the table, unfortunately I can't ship it anywhere, as they (rightfully so) don't want that much hardware out of their control.

So if we want to do the test we need to get someone else to donate the other stuff.

I am willing to travel a few hors to do the testing though.

slevy951

[quote]Originally posted by hplj123:
<strong>Sid,

If you can do that the YOUR THE MAN. please let me know how I can help.

Ken</strong><hr></blockquote>

Sid,

I'll make an offer of help....

If you want, since you have the equipment, I'll send you one of our strut braces to test with under these conditions...

1) You return the brace in sellable condition.
2) You have the option to purchase the brace when your done at retail price, if you'd like.
3) I get a copy of your methodology/results either way when your done.
4) I'll cover shipping both ways.

Contact me off list if you're interested in this offer.

If other brace manufacturers are looking at this and have a rigid brace they'd like to compare to ours, I'd be all for it.

SidViscous

To be perfectly honest yes it does apply to cars. It's called the scientific method and it is how all auto-engineers START making improvements. But to your point. At the end of the day yes it does need to be tested on the track. All part of the learning process, you then take the results and start over again, constantly refining the process.

To your point that is also why I said we could put an accelerometer in the car to measure the cornering force. So rather than having a driver say "it felt better/the same/slower" We will have actual data as to wether or not the car could take the corner faster. As to wether or not the added weight would reduce speed itself, I doubt it, but that is a seperate question. The question here is |Will a strut brace allow the 944 to take a corner faster/ or not| We can quantify that with numbers and not rely on subjective impressions.

The problem with proofing it with a driver is that all people are biased, and even subconsciously can spoil the results.

The only rall way to do that is with blind tests and multiple drivers which would cost allot more and take longer.

And that's not even factoring in drivers styles and how that comes into play.