Ot- Using deflection instead of calculation to determine beam strength.

Trboatworks · Mar 12, 2021

Another shop owner asked me a question as to if I knew a 'rule of thumb' for when a beam had displaced "too much" as a means of determining if over stressed in application.

I didn't as I usually run calculators for load case, beam type and material ME or just look at tables to determine beam size.

But.... he sent the following rule of thumb he found:

https://www.raeng.org.uk/publications/other/15-beam-deflection

Deflection Calculator for Square Tubing

What is the deal?
Rule of thumb a good field check but why get there- It seems the beam can be calculated before placed to meet loads.

The specific application was a queer one where the load case had surge load and twisting moments on beam so......
My rule of thumb in cases like this is to overbuild while relying on examples present for how this sort of widget is generally done.

Thanks

DDoug · Mar 12, 2021

If the beam returns to normal after the load is removed.

9100 · Mar 12, 2021

Stated somewhat differently, when the force plateaus and stays about constant with further deflection, you have reached the yield point and the maximum strength.

You can easily demonstrate this for yourself with a bolt through something like a steel plate and a torque wrench. I have seen old timers assemble engines by feel. When the torque you are feeling stops increasing, you are there.

Bill

crossthread · Mar 12, 2021

If you are going to use huge amounts of steel for a project like an office building or something then you want to know what is strong enough but the least you can get away with for cost purposes. I have tried to do the math on complex loads and it always seems to be a crap shoot. I am not however a structural engineer. For my projects, I just overbuild the crap out of it. The extra cost is made up for in peace of mind.

boslab · Mar 12, 2021

Should not exceed L/360 very thumby rule, designing by deflection is a bit retrospective
Mark

Trboatworks · Mar 12, 2021

9100 said:
Stated somewhat differently, when the force plateaus and stays about constant with further deflection, you have reached the yield point and the maximum strength.

Bill

Thanks that helps.

I think that is the problem I have with a specific deflection percentage as being a pass fail test.
It seems lots of factors such as section of beam and materials ME combine to change where deflection represents yield.

Scottl · Mar 12, 2021

digger doug said:
If the beam returns to normal after the load is removed.

Exactly. Elastic vs plastic deformation.

BT Fabrication · Mar 12, 2021

crossthread said:
If you are going to use huge amounts of steel for a project like an office building or something then you want to know what is strong enough but the least you can get away with for cost purposes. I have tried to do the math on complex loads and it always seems to be a crap shoot. I am not however a structural engineer. For my projects, I just overbuild the crap out of it. The extra cost is made up for in peace of mind.

well, overbuilt isn't always, ususally that is called safety factor, some are 3X some are larger then that.

DDoug · Mar 12, 2021

crossthread said:
If you are going to use huge amounts of steel for a project like an office building or something then you want to know what is strong enough but the least you can get away with for cost purposes. I have tried to do the math on complex loads and it always seems to be a crap shoot. I am not however a structural engineer. For my projects, I just overbuild the crap out of it. The extra cost is made up for in peace of mind.

Also, the catalogs don't show beams & columns in exact sizes you need, so rounding up adds to that safety factor.

TDegenhart · Mar 12, 2021

All this is done with a healthy dose of experience. Its one thing to do a simple beam test that is not going into use, its another to design a structure where human life is at risk. Finite element analysis can help, but in the end unknowns have to be accounted for. History is replete with examples of where the structure did not meet the usage requirements.

Tom

gustafson · Mar 12, 2021

With wood anyway there are deflection calcs as well as load calcs. Frequently a given beam can pass the absolute load but have more deflection than is desired. 1/360 is the 'won't crack a plaster ceiling' number, where 1/240 is OK for sheetrock with attic loads number.

With wood anyway, steel could be different

Trboatworks · Mar 12, 2021

gustafson said:
With wood anyway there are deflection calcs as well as load calcs. Frequently a given beam can pass the absolute load but have more deflection than is desired. 1/360 is the 'won't crack a plaster ceiling' number, where 1/240 is OK for sheetrock with attic loads number.

With wood anyway, steel could be different

This is a stainless square tube beam for controlling running rigging on a moderately large yacht- several ton load I guess.
Not my job- safety is there as a failure could make the main uncontrollable as well as the specter of a loaded line or spar taking out a crew member if one were caught on the loaded side during a failure.
I was on a similar size racing boat when a flying jibe blew up a turning block and locked up the rig with the boat over on its beam ends.
Middle of the night- ports open and I have never seen so much water flooding into a boat.
Whew- but no one hurt and they were able to clear the block and get the boat stood back up before it sank.

And YEARS before that for some weeks there was a large Skipjack which had done a similar jibe and roll with ports open and sank in the shallows on the northern bay.
She lay where she went down with sails full up just under the water till they got her raised- was quite a sight to see as you drove over the Bay Bridge..

The story was the working boat had been bought by a yachtsman who screwed up and sank her.

DanielG · Mar 12, 2021

Stress and deflection are related, but you can't determine if a beam is over-stressed just from the deflection. You need to know something about the geometry as well.

Take, for example, two beams with the same length. Beam A is made from 4"x4" square bar. Beam B is made from 1/2" x 8" flat bar, stood on edge. All else being equal, these two beams will have the same deflection. The stress in the flat bar, however, will be twice the stress in the square bar.

Things get even worse if you have a heavy section with an upstanding fin, like if you weld a piece of flat bar standing on end to the top flange of an I-beam. The added steel hardly makes a difference to the stiffness, but your maximum stress goes up significantly.

Nmbmxer · Mar 12, 2021

The only thing I've ever designed for deflection and not a stress limit are walkways. There is an SAE standard for that, but generally, walkways and operator platforms feel bouncy even if the stress is fine.

A good example of stiffness vs strength is to compare older US hydraulic cranes to modern European designs. The old US designs used 50ksi box booms that resulted in really stiff structures, later moving to 100ksi material but still a box that is limited by buckling and not a tensile stress. European cranes have much thinner booms made from 130ksi+ steel and formed into shapes that delay the onset of buckling to the point that you can reach the tensile stress limit. It's very un-nerving to watch them bending and deflecting so far even though you know both ways are equally safe.

DanielG · Mar 12, 2021

Nmbmxer said:
The only thing I've ever designed for deflection and not a stress limit are walkways. There is an SAE standard for that, but generally, walkways and operator platforms feel bouncy even if the stress is fine.

Natural frequency is also very important for not feeling 'bouncy'.

Flat roofs are also often deflection limited to avoid ponding issues.

9100 · Mar 12, 2021

About in the mid 50s a man who had worked for Curtis Wright told about building a test stand for a typical radial engine as were common on airliners. They used a standard nacelle with the motor mount attached to a stand of structural steel and "designed" by putting braces where they seemed appropriate. They could move the spinner up and down several inches by hand pressure. The kicker was that the deflection was in their massive steel frame, not in the really designed tubular motor mount.

Yes, Virginia, these guys wandering around with clip boards and slide rules do contribute to aircraft safety and performance.

Bill

Phil in Montana · Mar 12, 2021

When we use to inspect cranes we would use 1/360 as a guide and put the deflection down on the report...Phil

gustafson · Mar 13, 2021

Makes me wonder if that 1/360 means a damn thing or if that is the point where people say ' Is it me or is that bending?'

boslab · Mar 13, 2021

Phil in Montana said:
When we use to inspect cranes we would use 1/360 as a guide and put the deflection down on the report...Phil

I remember a load test on a 500 ton bridge crane, we had surveyors on the booms, and 1200 tons of dead weight, bridges are all generally positive camber but this thing flattened out and went slightly negative, I didn’t stay up there long, though the thing is still there in daily use, surprised how much it deflected but was told it was “normal” ( a new banned word apparently)
Mark

Illinoyance · Mar 13, 2021

1/360 comes from building and structural codes. Elevator machine beams are allowed 1/1666 deflection.

For the marine application OP described there is the possibility of loss of life or serious injury as well as damage to the vessel of several thousands of dollars. I think there needs to be a licensed engineer or naval architect involved. Why do members ask a bunch of machinists questions that should be referred to an engineering professional?

Ot- Using deflection instead of calculation to determine beam strength.

Diamond

Diamond

Diamond

Titanium

Titanium

Diamond

Diamond

Titanium

Diamond

Diamond

Diamond

Diamond

Stainless

Stainless

Stainless

Diamond

Stainless

Diamond

Titanium

Stainless

Similar threads