Tube Frame Design
 

If a guy would want to construct a tube frame chassis for a road vehicle, where could he get guidance for its design? I've looked on Amazon for books, I've googled, but haven't found much. There's books on suspensions and what not, but nothing really for the actual frame itself. One thing I have uncovered is Finite Element Analysis. I think the AutoCad Pro program can perform FEA. Is this what you need to calculate strength of a particular design? AutoCad Pro is darn expensive, though...

For a general overview I would recommend looking at dune-buggy frame plans.
Building your dune buggy. Custom building instructions for you!

You could take an existing design and alter the layout for the wheel size, ride height, and interior configuration you're aiming for, or just use it as a reference for how the structure needs to be laid out to handle the drive train and suspension element stresses while protecting the occupants.

If you want free Solid CAD try Alibre Xpress
http://www.alibre.com/xpress

There are also books about race car fabrication that will deal with tube frame design at least in one section. I am guessing that the books about building a "Locost 7" (that's a home-built design inspired by the old Lotus 7 car) would give you some basics, at least for that specific design.

This book looks like one of the "Locost" books:
Amazon.com: How to Build a Cheap Sports Car (Motorbooks Workshop): Keith Tanner: Books

-soD

Autodesk Student Community

There's free AutoCAD for students. Have no idea about the FEA.

Besides just the design of the chassis, the material selection is very important. What works for a race car with a 2 year life is very different than a street car that lasts 10 years and 100,000 miles.

I had it lucky I had my fellow teachers at the college here plus help from a local university for the information I wasn't able to find otherwise(structural engineering and the like). But if you are not going to college or working at one just showing up asking questions will probably not get you much help unless you can get someone there interested in your project. There are enough 'experts' on the internet that I would probably base most of my research on books and very little on web based information if you can't interest an engineer in helping. Some of the frames I have seen people build online are pretty scary.

Also don't ever use chromoly, structural steel is much better for road use, trust me. :D Using any tools like Autocad can give you a false sense of security if you don't have experience building things with it. It is one of the more annoying programs to use. I say just jump in and start building. When in doubt run over it with a truck or beat it with a sledge hammer and see what breaks. That usually gives pretty good results. But relying on computer tools to do your testing is nowhere near as good as testing out ideas yourself.

Why is structural steel a better choice than chromoly for road use?

"4130 Chromoly or ( Chromium Molybdenum ) has a 0.30% carbon content... thus you will have a higher tenancy for stress cracking or metal fatigue, specailly around the welds and transition where stress loads are high. Chromoly still is considered technically a "low carbon alloy steel." But 1020 steel ( Mild Steel ) is also a "low carbon steel" but it has a content that is 0.20%. The carbon makes the metal easier to weld, but the more carbon content makes it more brittle and thus lower the tensile strength of the metal. However, these metals both have a very high tensile strength."

Grabbed that from some random search result. It gives you the basic idea. A race car gets ran hard occasionally and retired after 2 years. The loads it has to handle are much higher than normal driving but the constant vibration and use of a street car will make it fatigue just as quick as racing frames. Unless you plan on replacing it every few years you will start seeing stress cracks show up all over it and after 5 years of use it will be so weak that any impact at all will probably rip the car apart.

The only bad part is you are probably going to double the weight of the frame you build by going to a structural steel frame. The good part is the frame will outlast the car. I guess everything is a tradeoff so it all comes down to what you want out of the car and how you plan on driving it. I am just trying to make sure you get all the information so you can make good decisions when building your ultimate car. It is a lot easier to build a car than most people think, it just takes a huge amount of time. Hopefully you will have lots of fun doing it :)

The classic on the subject is "Racing and sportscar chassis design" by Costin and Phipps. It's from the 60's (I think), but has absolutely everything you'll need to know about design, materials, suspension, you name it. Most libraries have it.

Dang. Another $100 book to put on the wish list. :D
Thanks, davidgrey50. That's what I was looking for. :thumbup:
I've also been wanting to get Hucho's book through inter-library loan again, so I guess I'll be going to the library tonight...

Coyote X, I have much respect for you after watching what you did with the Metro. I would do well to just jump in, as I do have the tendency to over analyze. But planning and designing will keep me occupied while I save my pennies for a welder. Speaking of which...do you prefer oxyacetylene or MIG? I learned oxyacetylene back in A&P school over 15 years ago, so I think I could dust off those skills fairly easily. Never tried MIG, but from what I understand it's not too tough to learn.

Keep in mind the metro was car #2 I tried building. The first one was a bit more rough :) It does take a bit of practice getting it right but the good thing is the metal is not that expensive really so it won't cost much to abandon a car and start over.

For welding I say use what you like. As long as it is a good weld it really doesn't make that much difference for most things. I used a gasless mig for 90% of my frame and a stick welder for the rest of it. I would have preferred to use a stick on all of it but the mig is so much easier to get a good weld with it was a matter of convenience to just use it.

Off the top of my head here is a list of stuff I have for metal work. I will probably miss something but it will be a good start.

Welder - lincoln pro-mig and lincoln ac buzzbox stick welder
cheap chop saw from Lowes
Angle Grinder came with it in a package deal
Tubing notcher from harbor freight with hole boring bits from Lowes
Bender from harbor freight
alignment magnets from harbor freight
hobart airforce plasma cutter
lots of jack stands probably 12 of them total. Useful for holding stuff up. Can also weld them in place to keep stuff steady.

That is about it and it has done a pretty good job for what I use it for.

Information on building a tubeframe
 

Tango Charlie,

If you don't know where to start, buy, beg, borrow or steal "build your own sports cars for 250 pounds" and build that one. It should take you about 2-3 years. This is coming from a guy who has built one (not for high mileage but for brutal performance on a budget - total cash outlay $7500) Mine's the one on the top of the wikipedia page! It's since been painted.
David grey50 is correct about "Racing and sportscar chassis design" by Costin and Phipps. It is a fantastic book and highly recommended - but unless your math is strong enough to crunch the numbers, build the seven replica first. I can't stress highly enough how much you will learn.
Then make a slippery body for it. Mine turned out a bit heavy at 650kg.
Coyote X is also correct in stating "It is a lot easier to build a car than most people think, it just takes a huge amount of time." There is lots of information
on the locost forums particularly if you ignore anyone "still in the planning stages" . Chromoly is unnecessary, expensive and your first car should be conservative enough in design to not require it. You can use it however, but I wouldn't on the first one, 1020 is much more forgiving. Chromoly will crack if the wrong filler material is used and/or allowed to cool too quickly. Right now I'm building a tandem seat with a 500cc engine, it was for the x-prize, but we aren't far enough along to be sure we'll be done by the date, and besides the $7500 entrance fee is half the budget for the car. If you can hold out till July I should be able to sell you a set of plans for $50.00.

best of luck

Stephen Flood

That's some car you've got there!

What's your MPG? :D

No idea, every time I drive it the accelerator seems to end up on the floor.
;-) (4age twin cam 1600cc stock internals)

Forget the oxy, ( unless you are using Chromoly, then it's a good idea) buy a mig.
(Miller, Hobart, Liquid air, lincoln.) it's easy to learn and about 10 times faster.

anybody15, welcome to ecomodder! You've done an awesome job with your Locost! What kind of engine did you use? I always thought a Locost with a Honda S2000 drivetrain would be fun around some twisty roads...
Actually, I do own Ron Champion's book. It's a great book if you want to build a Locost, but doesn't have any info on how to work out your own design, as that just isn't the point of the book. I'm wanting to build something I can commute to work in and get insane mileage; I've got some ideas that are begging to be proven.

Now a Locost Trike would be awesome-use the front end of the car, A Diesel like the one featured on the MAX design...the Max is getting over 70MPG as is, wonder what would happen minus another, say, 300lbs on the frame?

Tango Charlie,
The locost has a 4age corolla twin cam drivetrain (they were rear drive until 1987)
If you have a look at the X-prize pages there is a bloke entering a locost with a kotuba diesel engine, 60 odd mpg with aerodynamics of a brick. I am absolutely serious about
"build a locost first and drop a slippery body over it". With a small diesel you should be pushing 70mpg. Design your own is a great idea, but it is a STAGGERING amount of work. Don't get caught up it the idea rather than the task. You could spend 4 years designing something before you build anything - Don't. Hit the GO button. Make a light car ( and the locost is about as light as one can really go without moving to expensive composites), with lightness in mind you should be pushing 500kg. put a small/light engine in it (lupo diesel from europe?)

BTW the seven is for sale

Cheers

Stephen

For anyone that does want to get into design & doing finite element analysis and that sort of thing, there's this: OpenFEM for free. (I've never used it, just remembered it popping up in some other search, so caveat downloader :-))

Thanks, that's the title I was trying to think of!

-soD

I've heard lots of good things about Herb Adams' book Chassis Engineering. It's primarily oriented toward suspension theory and design but there is a frame design and construction section. I haven't read it but when I was in college on a FormulaSAE team the chassis/suspension guys pored over it for months (I was in the engine group).

The Formula SAE cars that use tube frames all use "chromoly" tubing of various mostly prescribed alloys (roll hoops have to be a certain alloy per the rules). They welded it using TIG with alloy-specific filler rod and used both pre- and post-heating. That's really the only way to avoid stress cracking long-term. Those cars were "thrown away" after 1 year of build/race and another year of test/train, but only the ones that were wrecked showed any frame damage. You can probably MIG chromoly but you would have to pre- and post-heat to get decent welds.

This would be a good place to start some reading.
Welcome to locostusa.com • Index page

Bookmarked. Thanks!
Hopefully the interlibrary loan for my copy of "Racing and Sportscar Chassis Design" will arrive before I have to head out to NE on business next week...

YOu could find a retired dune buggy, and cut it modify it to your need. MIG is faster.

Oh yeah and a lot of these are based on VW chassis.

"Tube" frames subdivide into simple beam types, such as the AC Cobra, or the Locust. The Locost has a "space" frame, using a 3-D truss instead of a pair of main tubes with some cross-connection and various extensions. The space frame is far lighter for the same stiffness.
Frames are primarily built up for stiffness, not strength. Thus, mild steel is often just as suitable as 4130 aircraft tubing, titanium or aluminum. What it lacks in ultimate strength goes into useful toughness in a crash.
For a great overview of the options, see "Structures" by J.E. Gordon - it is very readable, and not expensive on eBay.

There are some great software for tube bending and design,some are free.I would not try to reinvent the wheel,look at what others do and build off it.I would not use anything less then 1026 DOM or 4130 steel,you can't go wrong with 1 inch or better,using 4130 you can reduce the thickness for weight saving's but 1026 will be more practical.Look to JD squared for some great ,affordable and safe bending tools.Be careful using cheap bender's,I'm thinking about buying the model 32,hydraulic bender,1000 plus dollars and you can make headers,exhaust,chassis,motorcycle frames etc.Excuse my writing,I just like to get to the point.good luck.

For 4130 and other "chromoly" steels, you need to do a bunch of heat-normalization after welding. Mild steel, while heavier, is easier to work with.

This from several metal-workers and fabricators that I know.

-soD

If you are going the tube frame route, you might look at composite tubing. There is "Pultruded" fiberglass tubing in many shapes, round, square, I-beam, flat, angle, etc., which can be stronger and lighter than steel or aluminum, plus no exotic welding skills are need to join the tubes.

Fiberglass tubing
 

I just received a quote from a retailer who will sell small quantities of Fiberglass rectangular tubing. 40' of 2" x 4" for $869 including delivery, this is still too much. There is 184' of surplus 2" x 5" in Springfield Missouri for $225 but they will not ship.

The joints of any space frame structure . . .
 

. . . are very critical. With metal tubing, it is a simple exercise of fit and welding or brazing. With brazing, gusset plates may be needed depending on loading.

Designing composite joints is even more complex as it is not a homogeneous structure that is transferring the loads. The joints in your structure could very well be the weak links in your car.

Per the structural fiberglass manufacturers web site, these components can be mechanically connected. My plan for the production of the Auto X-Prize car line, was to quick connect the tubes with sheet metal and rivets, and then down the line the joints would be covered with fiberglass and or carbon fiber.
Welding is a craft, good welding is an Art. Welding a space frame car involves vertical and overhead welds, unless a means to rotate the frame is available, a car rotisserie. Fiberglass work is labor intensive, but the actual labor is low skill. Today I think I may try to use PVC fence rails to fab the chassis and then fiberglass over the joints with multilayers at the joints. I'll get some and try some sample joints....to be continued. :)

I just read your post and it's very interesting. Have a question. I'm not sure if I understand or not, but is it possible to do a frame with no welding?

thanks

One could do a Monocoque frame of folded sheet metal and rivet it together, strengthened with adhesive. Or Fiberglass of the same design as the sheet metal, or fiberglass tube with gusseted joints. Just to name a few possibilities.

I had to look up Monocoque, but what you're talking about sounds like just what I want. I like the first suggestion. It seems like an easy way to go for my first build. I'm going to have to look into it more. I really appreciate your post and it gives me some ideas.

I have one question... I looked at Harbor Freights and they have some rivet guns. Would a hand gun riveter be ok to use? This may be a dumb question but I'm just learning.

"Maybe", I have not had any experience with that HF tool, I am not thrilled with most of their stuff. There is a light, home made car on Ecomodder somewhere, that was build with flat sandwich composite panels. You might look up homemade aluminum aircraft construction, Aircraft Spruce company supplies a lot of materials for the home built aircraft community, and have a lot of experience with them. They have construction books, tools, and materials. For high quality rivet tools, I'd find out what you need, and go to Ebay or Craigslist for Good used over HF if they were going to be used a lot.

Search for Autospeed light weight car here at Ecomodder, the build series should come up.

I haven't looked on Craigslist yet, but I did read the thread you suggested. There were some other threads that were related and I'll read them too. Thanks so much for the help. This new information (for me) really makes it where I can see how things would work.

You're a good man :)

I've found a good EV body to build and would fit the use of rivets. Next thing I need to understand is what motor to get. I've read a number of threads, but they are mainly talking about something specific about a certain motor. I've read about forklift motors too.

So here's the question... If you're dirt poor, what motor would you use?

thanks a bunch...

For a light car I'd look at golf cart motors. You could use two, one geared low for acceleration from a stop, and one geared higher for higher speed, or use a CVT belt drive from a snowmobile.

Thank you so much for all the help. What would be easier.. the golf cart or the CVT thing? I think (but not sure) that I could find a golf cart before I could find a snowmobile. I haven't actually looked for a snowmobile because in the south, there are not many around :)

EDIT: I guess I should ask which is better... golf cart motor or CVT.