Material properties 3

[daak]

I'm running stress analyses for two types of equipment racks in which I'm given the following information about the materials used: one rack is made of 1" 4130 steel square tubing (mil-s-4130) and the other rack is made of 1" 1010 mild steel square tubing. I'm assuming that this is not enough information, as I would also need to know the type of treatment (ie. annealed...). Am I correct in this assumption or is there a reliable source to find this information? Thanks in advance!

 

[rb1957]

you're right, you need to know the heat treat of the steel (Annealed, Normalised, 1/4 hard, ...) then you'll need strength data from MIL-HDBK 5 (available on-line) or MMPDS-01 (maybe on-line) or maybe MIL-SPECs (available on-line)

 

[Sparweb]

Expect neither rack to be heat-treated in any way after welding. The welding rod must be known. ER70S maybe?
Was it welded by someone with AWS qualifications or cousin Eddie? Has anyone told you the wall thickness of tube used? Textbook information is useless until you know the rack's origin.

There is a "big picture" to consider when doing a structural analysis on something like an equipment rack. In some circles is referred to as the "burden of proof", but we're engineers and don't usually say it that way. Anyhow, if you assume the role of providing the "proof" to the Administrator (assumed to be the FAA in this case) then you must handle facts more than assumptions, and minimize the harm from possibly wrong assumptions where you can.
Trying to come up with an analysis for a multiply-redundant structure like a welded rack is difficult unless you want to go all-out with FEA. Even doing so, you cannot overcome the unknowns such as welding technique.

As I suggested (and so did the others) in the other thread you started, a structural load test is the preferred way to prove its strength. It eliminates (almost) all doubt.


Steven Fahey, CET

 

[daak]

I know that the welding is done iaw mil-w-8604A. The rod isn't specified on the drawing and I'm researching mil-w-8604A for this information. I know the tube wall thickness to be .060" and FEA is to be the method of analysis. Thanks for the "big picture" explanation. Yes, computer analysis is only as good as the input...

 

[tbuelna]

daak,

You cannot do a proper structural analysis until all of your weldment materials, heat treatments and weld joints have been specified in detail.

MIL-HDBK-5, sec. 8.2.2 gives allowables for the fusion welded section of mild and alloy steel. You will also need to apply the appropriate knock down factors for substandard weld conditions like undercuts, porosity, lack of penetration, etc. Unless the finished weldment is stress relieved, you should also assume there will be some residual stress within the parts due to the welding.

Hope that is helpful.
Terry

 

[daak]

Very insightful information, thanks Terry!

 

[RPstress]

Strictly, you can't do a proper stength analysis without more materials/joint info. The modulus of low carbon steel (and most other metals) is not much affected by thermomechnical history, so you can go ahead with material-linear FEA (details of joints, with undercuts as described by tbuelna will have a small effect, but can probably be ignored in terms of structural stiffness in a first/second pass).

At some point someone will have to decide on allowables. In the absence of any more information the only safe thing to do is use the minimum possible, which is probably annealed.

It is more complicated if you are sizing the part rather than just analysing it structurally. You can get a force distribution and then set thicknesses, etc., when allowables are known; however, very few real-world structures like racks are statically determinate. The alternative is to set some sizes based on known available material conditions, and specify material condition as well as sizes. This may well depend on post-weld heat treatment to regain quenched and tempered strengths, which will distort the part with this sort of low alloy steel. Advice from a metallurgist is advisable before putting too much time into it. An initial stick model with bars is easily adapted to new tube section properties and could be built and left with essentially guessed sections until more is known.