Gary_Ash

1932 Studebaker Indy car build

Recommended Posts

I've been laying out the cross members that will act as supports for the front and rear of the engine. The engine came equipped with a heavy (~.164" thick) Vee-shaped stamped steel front frame that has a rectangular steel/rubber/steel isolator under it. As the old isolator is now 77 years old and has been soaked in oil all those years, I ordered a new one. I also ordered the two pairs of rubber donuts that go at the rear under the bell housing, secured with 1/2-inch bolts. I need those in hand to figure out the height needed for the cross members. Isn't it nice to know that replacement parts can be ordered for engines this old!

I'm currently planning on using 1.5 x 3 x 0.12 inch wall type A500, grade B, rectangular tubing to make the cross members. I'll have to cut some pieces and weld the butted right angle joints. To prevent weld cracks from stress, I'll also add some 3x3 or 4x4 triangular gussets from 1/8th inch plate on front and rear surfaces of the butted joints. I'm figuring the lap welds of the gussets will prevent flexing of the butted joints. The engine, transmission, and accessories will weigh about 750 lbs. My calculation of the bending at the center of the cross members shows about .010-.015" for the case of both ends fixed, i.e. 1/4-inch mounting plates bolted through the chassis rails. Stress levels should be 3000-4000 psi, which is way below the yield strength of 46,000 psi of the steel.

The only thing making me pause at this point is that my cross member designs are not pretty. I thought about having the tubes bent, but bending rectangular tube this large is tough and the radius needed is too tight - the tubes would wrinkle and collapse. I could make curvy cross members by plasma cutting the front and rear faces and bending 3" wide strip to fit the contour. I'd be welding for a week and I don't think I could get it as strong as drawn tube. With the hood closed, the cross members won't show, and they'll still be hard to see with the hood open. Maybe someone else has some suggestions for alternate concepts or methods, though.

post-47871-143142806577_thumb.jpg

post-47871-143142806543_thumb.jpg

post-47871-14314280655_thumb.jpg

post-47871-143142806554_thumb.jpg

Edited by Gary_Ash (see edit history)

Share this post


Link to post
Share on other sites

Personally, I think you're over thinking things. The parts that aren't out front won't be a big deal in the long run... but if it really concerns you, I've always found that lightening holes we're a good addition. They break up the visual of large structural beams.

Share this post


Link to post
Share on other sites

From the structural point of view, you would be better to have single diagonal members from the chassis attachment to the lower horizontal member. That way you only have two knee joints, at more than 90°. Think of a truss - they don't have that inefficient rectangular shape. It will also be cheaper to make this way - fewer joints - and there will be fewer welds with the possibility of failure. The forces will be more axial with less bending moments in the joints.

The car manufacture would have bent something up. The 1939 Commander has an inverted U-shape front cross member with a "sag" in it, but not to the degree you appear to be looking at. The chassis on the Commander is an open channel as yours is and the cross member is rivetted to the top and bottom flanges as well as the web of the channel. You will need to use high strength friction bolts at least, with no threads in the connection; it should really be rivetted. (A car with bolted chassis like that would probably not pass the low volume vehicle safety inspection here.) You will be applying torque to the chassis with the mounting you propose and the chassis will probably deform in time and crack.

I think you are way short in your design loads. You have a heavy engine (= inertia) with a chassis bucking (up, down, left and right) and twisting under it. The engine will also rock quite a bit on that single front mount - the rear mounts will reduce the rocking, but the engine is long as well so will twist. (Were the racing engines on a single front mount?) The dynamic forces will be considerable - could they be 5x as much as the static force? The "rubber" mount will soften some of the movement but absorb little of the energy so the engine will still move relative to the chassis. In building base isolation we use "rubber" for springs and lead or steel plugs to absorb energy (i.e. like a shock absorber) and thus reduce movement between the ground and the building. The original cross member was substantial for a reason and not to just hold up the engine.

Share this post


Link to post
Share on other sites

Whtbaron and Spinneyhill: You guys continue to be very helpful and knowledgeable observers and constructive critics of my efforts, as well as others. I appreciate it. It isn't easy to find people who can discuss this stuff with knowledge and experience. I was looking for your input.

So, am I overthinking it? Yes, I am guilty, but that's part of the fun. I would make an exact replica of the 1931 car (the #37 car of Bob Valpey) but it isn't practical and I have some of my own ideas tempered by the fact that 83 years have passed since the car was created. We must have learned SOMETHING in the meantime.

And, Spinneyhill, I did try the diagonalized approach but I couldn't get it to wrap around the generator and other accessories unless the angles were already beyond 70 degrees, hence the rectangular approach. Also, the sedan engines sat much higher with respect to the frame rails than in the race cars.

You are correct: riveting would be the right way, as the engine cradle is currently formed and fastened. However, the race cars were made to be driven, crashed all too frequently, and disassembled to be repaired. Hence, the chassis is bolted together. These days, we do have Loctite compounds of blue and red types to lock things in place. Hardly any young engineers know the virtues of rivets, and modern cars do seem to live without them, truck frame rails particularly excepted. A rivet properly applied will never loosen and the parts will be clamped tightly enough that they never slide with respect to each other. One rivet-ignorant engineer I used to work with was afraid that a riveted joint would loosen with time and thermal cycling - I told him to stare out the window on the next airplane flight he took and watch to see if any wing rivets were rotating.

What has driven me to make these cross members different from the original cars is that the 1931-33 engines had "wings" cast on the blocks above the crankcase line, both in front and at the back of the block. My 1937 engines don't have those and depended on steel/rubber isolators at lower elevations and the stamped steel frames. The 1931-33 Indy cars had the engine block wings tied solidly to the frame rails, increasing the torsional stiffness and resistance to rocking motions. Of course, the vibrations were awful and the comfort was zero. I guess I am erring on the side of sedan comfort. On the other hand, the original cars have survived for 80 years without significant issues of frame or engine mount cracking, in spite of racing, crashing, road use, and who-knows-what over all that time. Heck, I'm already so old that joint failures in the car would occur long after any lawyer can come after me.

My calculations of the static loads resulted in stress levels that are <10% of the yield strength. If we now consider dynamic loads, it does add to the stress. My calculations show I might get as much as 300 lb-ft of torque from the engine, so let's pretend that I can get all 300 lb-ft transferred to the engine mounts at some point, perhaps even on a regular basis. I'll assume (incorrectly, of course) that the front and rear cross members will share equally in the torque transfer. So, I get 150 lb-ft of torque being restrained by two 1/2-inch bolts about 9 inches apart. This should generate 100 lbs of lift on one bolt and 100 pounds of compression on the other mounting pad. The front bolt spacings are similar. But, I won't double the stress on the cross members, and I'll stay well below my normal guideline of 1/3rd of the yield strength as the maximum load. Even cyclic loads of 1,000,000 cycles or more (i.e., infinite life) should be OK at these loading levels. Even if I add forces for bumps and jolts, I should still stay way below the yield strength. Of course, I don't want to be driving down the road at 70 mph and hear the "ping" of a weld failure as the car veers off the road. Do you think my 1.5x3x.12 rectangular tube has less moment and stiffness than the original sedan cross members? Mine are 26 and 31.5 inches wide. I don't have an old one to look at.

I do agree that this is an issue to be concerned about, both for me and anyone else building a speedster, etc. It's not just static loads, it's the repetitive flexing of parts that eventually cracks welds. Frequent inspections are critical on one-off machines. As a private pilot, I learned to walk around the plane before every flight looking for loose bolts, cracks in metal, and other signs of failure. In modern cars, people rarely open the hoods to check the oil or check the tire pressure. We've become complacent and lazy because the cars we drive every day are so safe and reliable. Just don't transfer that thinking to driving old cars or speedsters we build.

I will take your good advice and review the designs. Thank you!

Share this post


Link to post
Share on other sites

Another wildcat thought for you. My 1930 Dodge Brothers Eight has the front engine mounts formed from a plate bolted to the front of the engine under the timing gear cover. It is bent up at the ends to form bolting points to the chassis and to brackets attached to it. There is maybe 5 mm of "rubber" body mounting under the mount. The car is very smooth - you can balance a 20c piece on edge on the top of the engine while it is idling. The plate is 5 mm thick. The crankshaft (as denoted by the crank handle hole in the bottom of the radiator) is below the chassis.

I would imagine road roughness bumps would put significant loads on the engine mounts. The chassis moves upwards and downwards rapidly at speed so the forces on the engine could be significant. I would try to make some allowance for them, but I don't know how at the moment. With fairly stiff springs (e.g. for some attempt at racing handling) you will get forces and movements more than you might expect.

For transmitting shear, how does the cross-sectional area of steel of your proposed RHS compare with the original mounting cross member? I wonder if 3 mm wall thickness is too thin. What is the chassis made of? Is that 0.164" (about 4 mm)?

Could you improve the knee joints' bending moment capacity if you use the RHS with the long dimension vertical across the bottom and for the chassis mounts and laterally for the sub-vertical members? This will, however, reduce the longitudinal resistance to movement of the engine and transmission.

Have you checked deflections (strains) as well as stress? The design might turn out to be governed by deflections rather than stresses.

No doubt you are familiar with steel structures design code(s). I just thought they might be a useful reminder of what you need to consider.

Share this post


Link to post
Share on other sites

All good suggestions! Dwight's cross member image is interesting. I could imagine using 3/8"or 1/2" steel plate and having my local metal shop laser cut it and bend it to shape in one of their presses, then weld in the square or rectangular tube section and the gussets. But, it's hard to get stiffness out of a thin, flat plate. I've never seen what the 1937 Studebaker sedan cross members look like, so I can't guess at width, thickness, shape, etc. I've asked a friend to send me some photos. As in the design Dwight posted, I need to come straight out from the frame rails (0.164" thick mild steel) at least 2" to clear the lower flanges, then drop down quickly to get past the generator in front and the bell housing/clutch in the back.

When I did a simple analysis on the deflection and stress of the mounts, I had to assume that the mounts were simple, straight beams and ignored the vertical sections. I assumed the ends were not pinned [though they are bolted firmly to the side rails] and all the load is at the center of the beam. This gives simple calculations and some idea of what's going on. The stress and deflections were low enough that I'm not too concerned that the real cross members are much more complicated. Assuming that the end plates of the cross members are attached to an unyielding frame rail reduces the deflections at the center by a factor of 2.5. See a JPG of the calculations below for a couple of versions of rectangular tube dimensions.

To do any real analysis requires having some Finite Element Analysis software to get stress and strain. My TurboCAD 3D design software is pretty good, but there is no FEA plug-in yet. Where I used to work, they have SolidWorks with COSMOS FEA, an excellent package. But, it costs $10,000-$15,000 per seat or more. I just did a search and found the site for LISA, some free FEA software for up to 1400 nodes and only $99 for a full license for many nodes. See www.lisafea.com. It doesn't look like I could import a design into LISA, would have to create the model within the software. This could be an interesting diversion for a few weeks. Has anyone used LISA? Here's a sample of LISA output for deflection and stress of a simple cantilever beam.

Estimating dynamic loads like bumps, chassis twisting, cornering, etc., is very difficult. I got a copy of "Race Car Vehicle Dynamics" by William and Douglas Milliken (father and son), published by the SAE, ISBN 1-56091-526-9. See http://books.sae.org/r-146/. It's like a college textbook of race car design, 900 pages for $99. I concluded that dynamic loads of a few g's are common, but don't expect to see 10g and higher loads - the car would get launched into the air first.

Many years ago, I knew an old-time chief engineer who was responsible for the design of large high-vacuum chambers and complex mechanical systems. Back then, 3D CAD and FEA weren't available to anyone but places like Boeing and NASA, but he coped well. His favorite quote: "Steel is cheap, son. If in doubt, use more steel!" It's still good advice. Sometimes, it's just cheaper and faster to go put something together and see if it works, then modify as required. At this point, I'm really interested to see how much the chassis will go down on the springs when I put the engine and transmission in place.

post-47871-143142809027_thumb.jpg

post-47871-143142809018_thumb.jpg

Share this post


Link to post
Share on other sites

Hello Gary ,I am trying to reach you via your old e-mail adress but you do not answer.Is it still active or are you just upset about something?

Regards Thomas

Share this post


Link to post
Share on other sites

You're a way ahead of me with the computer design diagnostics. I tried to use autocad once and never got beyond making a straight line. I like the crossmember that Dwight posted, but I would also like to add more webbing to the outside wings. I was thinking about the later (late 40's early 50's) Chrysler motor mounts that were advertised as "floating power". Basically they mounted the engine with high connecting points to eliminate a lot of the vibration and wobble associated with inline motors. As mentioned above, the isolating rubber strip was quite narrow compared to modern motor mounts. Notice that this one is also built off center to resist the torque loads. Have to say it would be prettier if it was just stronger and not lop-sided. Hmmm.....looking at it again, the design may have had more to do with clearing the generator and crank pulley than torque loads.

post-47067-143142810781_thumb.jpg

Edited by whtbaron (see edit history)

Share this post


Link to post
Share on other sites

I have been trying to figure out some details of the engine mounting and exhaust routing. I got a drawing made for the 8-into-1 exhaust manifold using eight 90-degree bends in 1.5" dia. x .035 wall 304 stainless, a 1.5"x3.5"x28" long x.035" wall megaphone, and a 3/8" thick stainless flange. I just got a quote from one shop for more than $5000 to form the parts, machine the flange, polish the pieces, and weld it together - but final details could bring it even higher. I was blown away! I was thinking more like $500-$1000 so that I didn't have to cut, fit, and weld the pieces. I don't own a TIG welder for stainless, but at those prices I could afford to buy a very good TIG welder and spend some time learning how to use it. Is my reaction to the prices out of line?

By the way, I have a nice 3D PDF file (3 MB) of the engine assembly if you have Adobe Reader and want to see the engine details:

http://www.studegarage.com/images/indy/engine%20block%204-3D.pdf

post-47871-14314287757_thumb.jpg

post-47871-143142877551_thumb.jpg

post-47871-143142877562_thumb.jpg

Share this post


Link to post
Share on other sites

Have you got insulators between the carbs and exhaust manifold mounting plate you bolt them to?

What about if you made the manifolds of steel? I would imagine a not-small proportion of that price for stainless will be for polishing. It is also harder to work with than steel.

Share this post


Link to post
Share on other sites

$5000 does seem rather pricy for a basic 8 to 1 manifold, even with the tapered bit and being stainless. I was quoted $800 NZD for bending up the headers for my flathead 6 dodge motor and an extra $150 for lazer cutting the plate( the place I'm getting the stuff done dosn't weld them up) I'm only getting mine done in mild steel though.

Share this post


Link to post
Share on other sites

Gary,

This is beyond my level of expertise, but I would think you could purchase the parts (SS or mild steel) from a manufacturer like this (http://www.coneeng.com/pdf/cone%20list%20PDF.pdf) and have someone fit and weld the pieces for you. I would think semi-custom would be alot cheaper than full custom.

Dwight

Share this post


Link to post
Share on other sites

Yes, I thought it would have been cheaper based on the catalog prices of the company that quoted me for the fabrication. The megaphone was about $60, the eight bends were about $36 each. The raw stainless for the flange is less than $50, so about $400 for material - in stainless steel. They even offer their wholesale customers a 20% discount, so their real costs are quite a bit lower, say $200-$300 for everything. That leaves $4800 for their labor and profit to cut, polish, and weld. It should be about half that in mild steel. I can get the flange machined locally. If I use mild steel for everything, I can MIG it together myself. The biggest challenge is figuring out the shape of the hole where the bend in the 1.5" dia. tube intersects with the megaphone cone. My math skills are pretty good but that intersection is a mind blower. It's sort of worth trying to do it myself, if only to have a good model to give to someone else to do it better.

Share this post


Link to post
Share on other sites

The pdf drawing is pretty neat to play with! I haven't seen that before with pdf files. Xclnt.

Maybe I see the problem with your engine mount. It looks like you are supporting it below the vibration damper. If you support it on wings, like my Dodge or the Chrysler picture above, you remove this constraint. You also mention the racing engines had "wings" on them for the front mount. You already have a plate under the timing cover, why not turn it into "wings" to mount the engine?

If you look at plate 01-3 in the '34-'46 Chassis Parts Catalog, you will see the President and Commander Eight picture. The engine mount is pretty much behind the vibration damper. It is also made from an extension of the timing cover. Item 0101-1 (Plate) on Plate 01-8 may be able to be modified to make an engine mount for you, rather like the Studebaker "wings" you mention and the Dodge system. If you do it this way, you won't need to engineer your way out of a self-imposed problem = your proposed deep-drop cross member.

I can't recall what are you proposing for the radiator cross member. Rather than a deep-drop cross member, how about more rows in the radiator?

I now think your design will be governed by strains and fatigue, not stresses. I postulate the fatigue life is enhanced by keeping strains below a certain level. (I have been through the stress vs strain design route myself and got it wrong. Deflections were so great the things didn't work without stiffening to control strains.)

Another thing I notice in the chassis drawings in Section 15 of the chassis parts catalog is that the early chassis, Commander, Dictator and President, had the engine and radiator support cross member flared at the mounting ends, maybe to spread the stresses into a greater area of chassis.

Did the racing engines use a full length deep sump or a short sump to control oil surging away from the pump pick-up?

Edited by Spinneyhill (see edit history)

Share this post


Link to post
Share on other sites
The biggest challenge is figuring out the shape of the hole where the bend in the 1.5" dia. tube intersects with the megaphone cone. My math skills are pretty good but that intersection is a mind blower.

Surely your CAD program can draw a template for you?

Share this post


Link to post
Share on other sites

You can buy tube cutters & notchers that use hole saw blades to cut any angle and/or hole that you need. You set the tool up in a drill press to the right angle and right size hole saw and it cuts the right shape for the pipe intersection. Below is one supplier but there are quite a few out there.

http://www.medfordtools.com/tube_notching_101.html

http://www.harborfreight.com/pipe-tubing-notcher-42324.html

By the way, I also think the price you got is way too dear. Maybe try a local Hot Rod builder shop. If you do build them yourself out of stainless the cutting speed of the tube cutter must be very slow and you must use plenty of cutting oil/paste like Trefolex or one of those or it will burn the teeth off the blade.

Edited by DavidAU (see edit history)

Share this post


Link to post
Share on other sites

Yes, it looks like the tubing notcher is the way to go, and the Harbor Freight price is right! I won't need to worry about the shape of the holes, the notcher will make it just right. I can buy the eight bends and megaphone cone in stainless steel, plus a blank for the flange, at less than $400; the same in mild steel is $200-$250. I think I'll buy the mild steel parts, then I can weld it together with my MIG welder. This is one case where a TIG would sure make a better looking weld when it counts. It will be a little tricky to clamp the cone, but not too bad.

Spinneyhill, I did try to buy the front and rear engine mount plates as used on the 1929-32 engines, but couldn't locate a set at a reasonable price, e.g. buy an entire engine and ship it across the country. I'm having to make some compromises to keep the project moving. I don't want to run out of life or money before getting to drive the car. I agree that fatigue from flexing will be a issue in avoiding weld cracks or failure of chassis parts. I'm trying to stay below 1/3 of yield stress for static loads to allow for flex stresses that will still stay below yield. Look at the photos back in my post #76 in this thread to see the 1-1/4" dia x 1/8" wall tube I'm using for the radiator support. There are two bosses welded to the tube to hold the vertical brackets for the radiator shell and radiator. As with all of the rest of the crossmembers, the tube is welded to 1/4" thick end plates with three bolts per side for attaching to the frame rails. The radiator will be 4-core, 3-1/2" thick (more than stock), and it will use an electric puller fan since there is not enough clearance under the hood to spin one with the fan belt. The original cars had no fans. I may need to keep the fan mount and pulley just to have a way to adjust fan belt tension. I think the engine oil sump used on the engines under development in 1934 was the standard one. Of course, Studebaker always painted the insides of the block with red Glyptal, even on sedans and trucks, to eliminate casting porosity and improve oil shedding.

Share this post


Link to post
Share on other sites
Spinneyhill, I did try to buy the front and rear engine mount plates as used on the 1929-32 engines, but couldn't locate a set at a reasonable price, e.g. buy an entire engine and ship it across the country.

Can you get the drawings from the museum (or measure up a pair) and make some?

I see here http://1.bp.blogspot.com/_UlK0t3FwMY0/TNTVHEhaINI/AAAAAAAABJo/VkVGIKTartA/s1600/commander_8_engine.jpg the types of mounts you probably mean. The Dodge system is very similar, except the Dodge has a cast mounting "plate" on the rear. If you start with flat plate is should be easy enough to bend up a mounting to suit

Edited by Spinneyhill (see edit history)

Share this post


Link to post
Share on other sites

I'm a stubborn kind of guy, but having actually welded the rectangular tube mounts together, it turns out they won't fit and can't be made to work. Too much interference with the spring shackles. And, yes, they were UGLY! Anyone want some 1.5"x3" cross members strong enough for an M1 tank engine?

So, back to the (electronic) drawing board to model up some flat plates like they were going to use in 1934. I have the photos of the development engines then and the manufacturing drawings for the engine block I am using, so I have most - but not all - of the critical dimension information. I may be close enough to let the shop laser cut and bend the plates, though. The "wings" have to be short enough to allow the engine to be lowered into the frame rails. I'll need some other small mounts to couple the frame rails to the engine plates, as they did back then. I'm going to have to pull the harmonic balancer and the front cam drive cover to look inside, a task I have been delaying for a long time. The front plate has to have holes to provide space for the cam drive gears and the front of the crankshaft, etc. The front plate will be flat, 0.164" thick if I stick to what's there now, 0.25" thick if there is room to move the cover forward and not hit the harmonic balancer and crank pulley. The back plate can be 0.150" (9 gauge) if I can't move the bell housing backwards, 3/16" (0.188") if I can slip it in. I don't want the block oscillating front to back. Anyway, here are drawings so far with some photos from 1934 of the original engine. My mounts are a little different because the factory was probably trying to adapt to the side mounts that were there for the old 337 cubic inch engines used is 1931-33. I may have my crankshaft centerline at just a little different height. Once I get all the drawings done and am sure it will all fit, it's off to the sheet metal shop again.

The original rear plate had small stiffeners riveted to the main plate. I don't have the right stuff for putting good heads on 1/4" shank steel rivets, so I'll use 5/16" bolts with red Loctite to hold them tight in spite of the vibrations. Anyone have experience with blue versus red Loctite? The red stuff seems to be "forever"; the blue can be broken loose with a strong wrench.

Oh, the big notch in the front plate is to leave room for the generator clamp band. The original cars didn't have generators or batteries, just a shaft running back to the side-mounted water pump. I can't see how they planned to adjust fan belt tension on those engines - and there was no fan!

post-47871-14314290102_thumb.jpg

post-47871-143142900943_thumb.jpg

post-47871-143142900958_thumb.jpg

post-47871-143142900966_thumb.jpg

post-47871-143142900988_thumb.jpg

Edited by Gary_Ash (see edit history)

Share this post


Link to post
Share on other sites

That looks more like it. Pity about the "wasted" effort. I noticed on the Studebaker thread about an inherited 1930 FH President, the front engine mount (similar to this) had a top flange (bent backwards) for strength. No doubt the rear mount is similar.

Share this post


Link to post
Share on other sites

I saw the thread about the 1930 FH coupe. That would be a sweet ride, could be a nice "driver". I sure don't NEED another car.

The rear mounts on those big 337 cubic inch engines were mostly cast into the block. There was a bent U bracket with a stiffener plate that joined to the Indy car frame rails. Here are some photos of the green #37 car showing the rear mounts and some shots of the white #34 car when it was being restored for Brooks Stevens many years ago. Whatever front mount those engines had in sedans was replaced by a thick, flat plate for the Indy cars. Here is a shot I took of the front mount on #37 while lying on my back under the car. You can see the nice weldment that mates to the flat plate. I'll need to make something like this, too, as well as U-shaped ones for the rear. The Indy car engines were solidly mounted, no rubber. The early sedans with the 337 engine, like that 1930 FH, had little, if any, rubber.

post-47871-143142902492_thumb.jpg

post-47871-143142902478_thumb.jpg

post-47871-143142902483_thumb.jpg

post-47871-143142902487_thumb.jpg

Share this post


Link to post
Share on other sites

Xclnt. As they say around here, "you're onto it!" The 8-cyl. motors don't need much rubber, they are smooth enough to bolt directly to the chassis. 1930 Dodge is, anyway, although only a 220 cu. in. engine.

Share this post


Link to post
Share on other sites

Just catching up. I just had a thought - those flat plate mounts you show a couple of posts ago probably don't offer much control of longitudinal movement. How are you getting on with this challenging project?

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now