Designing steel frames and weldments

A. Hi John,

As Bob Dylan tells us, you don't need a weatherman to know which way the wind blows. You don't need a physician to deal with a small splinter or a cold. And so it can sometimes be with designing: almost everyone has slapped together a few pieces of lumber with screws or nails to make a dog house or a brace for a sagging pipe or a storage stand for their firewood; it doesn't take an engineer.

But when you are talking about steel plate thicknesses, weldment integrity, and loads of 55,000 pounds, it sounds like you are solidly into an area where the design should be done by an engineer. One of the reasons is that an untrained eye may not even see an unexpected point of failure.

Your terminology isn't right; 'psi' is a measure of the stress on, or strength of, a material per unit area. Are you trying to say that the load on this weldment is 55,000 lbs. tension? Engineers ALWAYS apply a 'factor of safety' as well, they never design to the yield strength of a material. The weldment may fail at the welds rather than within one of the plates anyway, and designing tension welds is a very tricky business indeed. In fact, it is prohibited to design many structures (like cranes) with welds in tension. There are other factors to consider like L/r ratio as well. Please call in a local engineer: it is illegal to practice engineering without a license; and for an engineer to try to do it by 'remote control' over the internet, with a poor understanding of the situation, would be just as bad and probably just as illegal :-)

A. I realize that skilled mechanics sometimes just piece together non-critical weldments based on their experience and judgement, Harry, and I'd be wasting my breath trying to condemn something that is so widely practiced. But you can't proceed to engineer a load-carrying device via internet advice. It's no different than asking for internet advice on how long to make the incision for an appendectomy. An engineer or architect spends 2 years in engineering school just in "analyzing the forces" before he can take his first introductory course in structures or design of machine elements. Even in a simple device there are many potential modes of mechanical failure, just as there are many ways the patient can get sicker or die during or because of surgery :-)

As one quick example, those four square tubes can probably support a lot of weight as long as it's purely vertical (normal); but with a wall thickness of only .045" they may not have much resistance to "racking" ("pancaking", "accordion-ing") from a side load. And once they've pancaked, now the thickness of your beams is .090" rather than 1.0". Suppose the purpose of your frame is to support one wheel of a heavy truck, and you calculate that it's strong enough... but then as the truck rolls onto your frame the driver jams on the brakes and "racks" those tubes?

The first thing that mechanics and welders should probably ask themselves is: "What is the very worst thing that can happen if this thing fails?". If the consequence could be serious, STOP until an engineer can review it. If you are sure that the consequences of failure are insignificant, enjoy!

A. Hi Marino. Type 304 stainless steel is readily available and has a fairly high linear coefficient of thermal expansion (higher than plain steel and most stainless steels), although not as high as aluminum.

But getting affordable stainless steel studs that are as strong as hardened steel studs will be a problem. If cost is no object, Ken Vlach discusses high strength stainless steel fasteners in letter 38035 "Galvanic corrosion of alloy steel fasteners in stainless" -- but cutting threads in exotic stainless steels may be difficult (they would probably have to be ground). May I ask why you want to use stainless steel rather than hardened steel? Good luck.

Regards,