Frame torsion testing underway

[69427]

Yeah, I spent some time bending stuff to see where I'm at presently. I was having serious issues with the anchor point (my carlift runner) bending, throwing off my "datum" point, as it were. I made some changes, along with adding some additional material on the runner to distribute the force over a larger area of the runner surface. Seemed to help.
I then placed my dial indicator at the frame area near the batwing mounts (which I'm using as the frame anchor attachments). I was getting some bending of the mounts during the torsion inputs. I suspect my mounts might be a bit more susceptible to bending versus the factory batwing mounts due to my having to narrow the batwing width to clear the C4 bearing carrier assembly, which then required somewhat longer and more angled mounts. I welded in some additional gussets in the mounts, which then reduced the frame movement in that area to about four thousandths of an inch. Small enough for now, and ignoring this error will give conservative readings on the measurements going forward .
I put my dial indicator under the rear (pinion) crossmember and then applied the 2000 ft-lbs of torque up front. I got 150 thousandths downward movement on the frame. At 26.5 inches from the frame centerline I got a calculated angle of .32 degrees, yielding a rate of 6250 ft-lbs per degree. Given how flimsy the rear kickup area appeared when I first got the frame, I can live with this number.
I next measured the delta movement of the area between the pinion crossmember and the trans crossmember. I calculated this angle at .24 degrees, giving me a torsion reading of 8333 ft-lbs per degree. I then removed the torsion tube I have mounted between the trans and pinion crossmembers. After applying the load again, I ended up measuring and calculating a reading of 6978 ft-lbs per degree, a reduction of 1355 ft-lbs per degree (or about 16%). I checked my calculations for the torsion tube (using the formula in my "How To Make Your Car Handle" book), and got a theoretical torsion rate for the tube at 1316 lbs per degree. I had to do several conversions to convert pounds per inch into pounds per degree, so I hope I did it right. But regardless, I'm still reasonably happy with this section of the frame.
The front section of the frame, from the trans crossmember to the suspension crossmember, is the most difficult portion to stiffen up. This segment length is roughly twice the length of the center and kickup sections, with the engine and transmission getting in the way of installing efficient stiffening structures. I measured a deflection (relative to the trans crossmember) of 370 thousandths. This calculated out to .8 degrees, and a result of 2500 ft-lbs per degree. For the same unit stiffness I would expect this section to be flimsier due to its longer length, but I'm still not happy about it. I dropped the engine and transmission back onto the frame a couple days ago to study where there's available real estate in the front half of the wheelbase to possibly add more crossmember/struts/gussets to stiffen this up some more. Given that these frames are essentially three torsion sections end to end, the calculations give me an overall reading of just under 1500 ft-lbs per degree. Stiffening up any of the three sections will yield a higher overall number, but given these first run numbers, improvements to the (long) front section should yield higher percentages of improvement to the overall number. The problem is that this section is the toughest to improve, given the space restrictions.
An interesting, but frustrating aspect of this has been the difficulty of trying to determine weak "nodes" in the frame. I put a couple laser levels on different portions of the frame and aimed them towards other areas of the frame to try to visually see the twist of one portion of the frame versus another. In most, if not all, instances, I could see the overall frame move or twist a bit as I applied the torque, but the lasers would continue to point to the same point on the frame. I've essentially got no usable information from trying this particular test.
Still got a couple things I want to do to the front of the frame, along with some possible measurement techniques to try to streamline the readings, and hopefully reducing any errors in these measurements.
More to follow during the week.

 

[Bowtie-racing]

This is really intresting test. Keep up the good work!!

 

[69427]

Played around a little last week. While staring at the frame and trying to figure out my next move, I decided to do some simple welding to make some productive use of my time while I'm out in the garage. I put in some boxed gussets at the four bends between the front bodymounts and the motor mount horns. Pretty straightforward other than making allowances for the fuel and brake lines to continue to be packaged somewhat in their original locations.
Went back and did some more twisting on the frame trying to look for areas that I can still play with, and made a couple measurements with my dial indicator while I had things set up. I didn't expect much change from my previous measurements due to the bend gussets being more suited to beam strength improvement rather than any torsion improvement. The front sectional (forward of the trans crossmember) torsion change was four percent (2600 vs 2500 lb-ft/deg), which then translated to a wheelbase length improvement of 2 1/2 percent (1517 vs 1483 lb-ft/deg). (I've still got a little distortion in my anchor area, which slews these numbers downward, but it's not more than just a couple percent.)
Outside of the gussets, I've been spending most of my time lately machining on the custom aluminum bearing carriers/caliper brackets I'm making to allow use of the Wilwood calipers I'm using. (These calipers package better than the stock C4 sliding calipers.)
I'll update things again when I get back on the frame.

 

[TimAT]

Looks like you're gaining on the final answers. I did come up with a thought (difficult for me, I know) Right now you're looking at a bare frame- I mean the body is off. Will the effect of the body being mounted add any rigidity. I know the bodies are not very stiff alone, but add the body and structure of it and then bolt it all to the frame. Will it make any difference?

 

[69427]

Looks like you're gaining on the final answers. I did come up with a thought (difficult for me, I know) Right now you're looking at a bare frame- I mean the body is off. Will the effect of the body being mounted add any rigidity. I know the bodies are not very stiff alone, but add the body and structure of it and then bolt it all to the frame. Will it make any difference?

Yeah, there ought to be some undefinable improvement in the torsion rating. The bodywork has some value of torsional rigidity to it, and when bolted to the frame, the car can essentially can be modeled as two torsion bars in parallel, which would be somewhat, I say somewhat, in the ballpark of the sum of the two torsion ratings.
Been doing a little more playing around. I (temporarily) welded in a couple scrap lengths of square tubing (1") between the trans crossmember (near the exhaust passthrough holes) angled forward to slightly rearward of the motor mount horns. This added a bit more rigidity (approximately 100 lb-ft/deg) to the front portion of the frame. Once I get the body back on, I'll entertain adding some permanent struts here, as right now it's a bit difficult to determine how much floor clearance I'll have with the body on.
My continuing frustration is the weakness of the batwing mount area. It continues to look like the frame is pivoting/flexing at that point (contributing to lowering the torsion numbers of the frame). Given all the suspension stuff moving around in that area (primarily the knuckles and toe rods), I may have to wait until I have the suspension bolted on before I can determine how much clearance I have to reinforce that area. I'll be out of luck doing much torsion measuring at that point, but that's life.

 

[BBShark]

When you divided the frame into 3 sections, were you measuring the torsion of the individual sections? When you say that the diff pinion crossmember was down .150, where was the torsional load applied and where was the rear of the frame restrained? Also, did the restraining method allow the rear of the frame to twist without adding stiffness of it own (for instance if a rigid structure was clamped to the frame)?

 

[69427]

When you divided the frame into 3 sections, were you measuring the torsion of the individual sections? Yes, in an indirect way. I measured the deflection (with my dial indicator) at the pinion crossmember, the trans crossmember, and the front axle crossmember. I then did some 8th grade geometry calculations to convert these deflections to degrees, and then factored these into the 2000 number to get the torsion results for each section. When you say that the diff pinion crossmember was down .150, where was the torsional load applied and where was the rear of the frame restrained? I've got a seven foot moment arm (2.75 inch square tubing) bolted across the front axle centerline where I apply the torque. A combination of the arm weight, my truck's spare tire, and me at this distance gives me 2000 lb-ft of torque. The rear of the frame was restrained by the batwing mounts, as this is the most rearward point that the suspension loads will input force.
Also, did the restraining method allow the rear of the frame to twist without adding stiffness of it own (for instance if a rigid structure was clamped to the frame)?

Yes. The batwing mounts are bolted to a 5X5 inch angle iron piece. This allows me the ability to clamp (rather than weld) the bottom of the angle iron to the carlift runner. The batwing mounts are allowed (wisely or unwisely) to rotate around their anchor mounts. When the mounts rotate, it gives the effect of lower torsional readings (more deflection). This persuaded me not to weld a small bead between the batwing mount and the anchor, as I thought this might give false readings, as the actual batwing is bolted to the mounts, and can rotate slightly under high loads.
I'll try to take a picture and post it. I finally figured out a workaround to my Photobucket problem of late, and can access my pictures again.

 

[BBShark]

VetteMOD has image hosting. It's easier to use than Photobucket, works better and it's free.

.

 

[69427]

Just a couple quick pictures before I crash out for the evening. The first shows the two temporary angled struts I welded in to see if there was any improvement. It helped a bit, as I described earlier, so I'll entertain permanent struts once the body is back on so I know how much room I have to put some bigger elements there.

This photo of the left rear batwing mount area shows how the frame is bolted to a piece of 5 inch angle iron, which is then clamped to the runner of my car hoist. During tests I'm seeing some twisting of the frame runner area above the batwing mounts. I'm going try to add a crossmember across the batwing area (if I can) to try to resist this twist. The room is very tight under the car in this area, so I don't know how it's going to go.