Torsional Rigidity Testing, 280Z

[gnosez]

Here's the link to the photos I had found in a roll of undeveloped 35mm B&W film from 1974 showing the body prep work on the BSR 260Z (gnose):

 

http://s160.photobucket.com/user/psanders240/media/BSR-260Z-IMSA/b1505291-9291-4ce7-98b0-6f829b79ff6e_zpsb9179a63.jpg.html?sort=3&o=0

[z240]

Wow, those pictures are a treasure of a time capsule and a lesson in strengthing the chassis.  Thanks for posting that.

 

My favorite feature is the camber/caster "adjusters" at the top of the front and rear struts. Big a$$ washer big enough to sandwich the entire strut hole!  They didn't actualy run like that did they? 

Edited February 11, 2014 by z240

[Chris Duncan]

Additional torsional rigidity testing.

 

Same vehicle with an added roll cage

 

1.5" x .125 wall DOM tubing, some 1.5" x .095"

 

Six point roll cage with an additional eight body-to-cage gussets of .095 sheet metal. At middle and top of A-pillar, and middle and top of B-pillar (back of door)

 

2400 pounds per degree, so about a 20% gain.

 

While doing the torsion test also measured between the front strut towers where a conventional strut tower brace would normally go. Zero measurable movement.

 

 

[1vicissitude]

Pics of the cage?

 

Anything forward of the firewall? I am guessing of course, but I assume a lot of the torsional movement is happening towards the front of the cabin and after the firewall.

[mattster03]

Wow, this thread is impressive...

[Chris Duncan]

Pics of the cage?

 

Anything forward of the firewall? I am guessing of course, but I assume a lot of the torsional movement is happening towards the front of the cabin and after the firewall.

 

It's a very standard 6 point cage, X's in the doors, 45's to the rear strut tops. Diagonal behind the seats. Nothing in front of the firewall yet. But it will end up with 2 more tubes running from the cage to the front strut tops.  I would assume the cage has moved the flex away from the center area to a certain extent.

Edited June 4, 2014 by Chris Duncan

[1vicissitude]

It's a very standard 6 point cage, X's in the doors, 45's to the rear strut tops. Diagonal behind the seats. Nothing in front of the firewall yet. But it will end up with 2 more tubes running from the cage to the front strut tops.  I would assume the cage has moved the flex away from the center area.

 

Is there a bar across the rear strut towers as well? I can't help but think that would make a notable difference.

[JMortensen]

While doing the torsion test also measured between the front strut towers where a conventional strut tower brace would normally go. Zero measurable movement.

This comment here set my brain in motion. Sorry if I rain on your parade here, not my intent, but I really think you've got some problems with your testing.

 

Seems to me that your test rig reinforces the front strut towers from twisting, and it actually might do that job quite effectively. You've got the lower control arm stand in and the strut stand in welded together and bolted to what looks like a 2x6 presumably pretty heavy walled piece of steel. That assembly is fighting the torsion that you're putting in, and I would think that it is adding a lot of stiffness compared to stock. Add to this the idea that you put the torque into the lower frame rail in the back, not the strut tower there, and your test is more of the lower frame rails and rocker's ability to resist torsion when aided by the front reinforcement.

 

I dunno. Maybe I'm wrong, but the more I think about this the more I think you need the actual suspension members in there and you need to put loads into them, not the frame rails themselves. You're putting torque into the chassis, but you've also reinforced it at the same time. My suspicion is that you'd end up with MUCH lower numbers with the front end tested at the hubs.

 

Just looking at the "Think Fast" book again, he describes mounting the chassis via the hubs, and says: "The wheel mount blades should be very stiff, but only in the vertical direction. They should be flexible in all other directions so that the fixture does not artificially stiffen the car." Your front and rear setups look rigid in all but beam.

 

I think in order to do this test and get really accurate results the suspension will have to be all metal to metal bushings. Your thread had me considering doing something similar, but my plan was to do it through the hubs. I have all monoballs and heims joints except stock front ball joints. I think even they would have to go in order to make this sort of testing accurate.

Edited June 3, 2014 by JMortensen

[Chris Duncan]

This comment here set my brain in motion. Sorry if I rain on your parade here, not my intent, but I really think you've got some problems with your testing.

 

Seems to me that your test rig reinforces the front strut towers from twisting, and it actually might do that job quite effectively. You've got the lower control arm stand in and the strut stand in welded together and bolted to what looks like a 2x6 presumably pretty heavy walled piece of steel. That assembly is fighting the torsion that you're putting in, and I would think that it is adding a lot of stiffness compared to stock. Add to this the idea that you put the torque into the lower frame rail in the back, not the strut tower there, and your test is more of the lower frame rails and rocker's ability to resist torsion when aided by the front reinforcement.

 

I dunno. Maybe I'm wrong, but the more I think about this the more I think you need the actual suspension members in there and you need to put loads into them, not the frame rails themselves. You're putting torque into the chassis, but you've also reinforced it at the same time. My suspicion is that you'd end up with MUCH lower numbers with the front end tested at the hubs.

 

Just looking at the "Think Fast" book again, he describes mounting the chassis via the hubs, and says: "The wheel mount blades should be very stiff, but only in the vertical direction. They should be flexible in all other directions so that the fixture does not artificially stiffen the car." Your front and rear setups look rigid in all but beam.

 

I think in order to do this test and get really accurate results the suspension will have to be all metal to metal bushings. Your thread had me considering doing something similar, but my plan was to do it through the hubs. I have all monoballs and heims joints except stock front ball joints. I think even they would have to go in order to make this sort of testing accurate.

 

Looking at what you're saying I think you are somewhat correct.

 

The front stand-ins are solildly bolted to the 2" x 4" x 1/8" testing beam, so that aspect is not going to simulate the actual parts. However the way the strut and lower arm stand-ins are attached to each other at the lower outer ball joint position is set up to flex, so that is closer to the actual set-up. So I think the torsional rigidity test is somewhat accurate and the front strut tower brace test is less accurate. So to improve this set up one of the testing beam attachement points would have to have a sliding feature so that the lower ball joint positions are not rigidly attached to each other.

 

In addition to increase accuracy the rear could be set up with similar strut and lower arm stand-ins. I had origially started to do this but in the interest of time decided the accuracy for the intended purpose would be enough with the direct bolt up to the chassis.

 

Chumpcar penalizes metal to metal bushings and to be 100% accurate you would also need solid strut top mounts. Going in I didn't think this would have full accuracy, what I tried to emphasize was uniform repeatability. Trying to apply the same test over and over to observe the results of different changes. And I think that was achieved enough for the intended purpose. The seam welding does not have enough return for the time and expense it takes and the roll cage does add a significant amount of rigidity especially with the added cage to body gussets. I believe it was also indicated by someone's reply that the overall rigidity number was close to what others have seen in the past.

 

One thing that was discovered that I haven't seen anyone else do is the method to load the beam, this makes quick repeatable testing much easier. Of course that doesn't mean it's never been done, just that nobody has put it out there.

Edited June 4, 2014 by Chris Duncan