Deflection in beams

[KieranSean]

What could be the causes for a difference in measured and calculated values of deflection, other than human error, material impurities etc?

 

[IRstuff]

How big a difference? Small errors are par for the course.

How well does your model correspond to the actual measurement conditions, loading conditions, and the actual beam properties?



TTFN

FAQ731-376

 

[KieranSean]

I've used macauleys method to calculate a value for deflection and in doing so i have only assumed the value for youngs modulus.
Ive also ignored the fact that it is a UDL as i zeroed the DTI prior to loading.
My measured value came out at 0.07mm and my calculated at 0.91mm, i have gone over my method a hundred times and still obtain a similar result.

 

[dvd]

end fixity is one cause.

 

[KieranSean]

Would end fixity still affect it if the supports are positioned in from the ends?

 

[dvd]

trying to solve a frame with beam equations comes to mind also.

fixity in general will affect your results. Knowing how to apply actual restraint conditions can be difficult.

You should post a sketch of your frame/beam.

 

[kelowna]

Is the way you are measuring OK? 0.07 mm is almost negligible, not many instruments will measure that correctly, even 0.91 is very little. The accuracy of your measuring instrument could account for the difference.

 

[FeX32]

multiply your force by 100 and redo everything you did.


Fe

 

[GregLocock]

A photo of your test setup would replace a million words.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[jte]

Is this a homework / lab question or is there a real-world problem here?

jt

 

[hokie66]

Amen to Greg's suggestion. Of all the structural materials, steel has the most consistent modulus of inertia. So, for a pure bending exercise, measured deflections should match computed. Support conditions can dramatically differ from assumed, and for short spans, members tend to act like tied arches rather than beams.

 

[drawoh]

What could be the causes for a difference in measured and calculated values of deflection, other than human error, material impurities etc?

Double integration method is only accurate for small deflections. One of my mechanics of materials books states that it is good up to deflections of 10% of beam length. How long is your beam?

Also, how accurate is the thickness of your beam? Your beam stiffness and deflection are a function of thickness cubed. Any thickness error will be multiplied. Beam length is another value that gets cubed.

JHG

 

[rb1957]

how ideal are the supports ? i imagine you analyzed them as pinned (or fully fixed) ? how accurate is your measurment ? (0.07mm is pretty darn small) ...

you're looking as max deflection, yes? instead of repeating the same calc (potentially repeating the same error), what does Roark (or some other handbook) say the deflection should be ??

 

[electricpete]

If your measured was more than your actual, I’d ask if this were a short stubby beam where shear deformation needs to be included. But your actual is more than calculated.

More knowledgeable people than me suggested to pay attention to boundary condition assumptions – makes sense to me.

Some possible ways to explore further:
1 – take multiple measurements to see if the shape is as predicted by beam theory.... and examine any deviations as a clue to the behavior.

2 – Post a more detailed description of the physcial problem here. Then people can provide comments on validity of assumptions for that problem and also possibly double check the conclusion.


=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[desertfox]

Hi Kieran88

Well more information on your beam set up might help getting a better understanding of your problem and ultimately a better answer.
The difference between a point load and a uniformly distributed load, would result in a deflection of 1.6 times the deflection of a uniformly distributed load to get to the deflection of the equivalent point loaded beam, everything else being equal.
This clearly doesn't account for your difference in deflections so we need to look at other things.

desertfox

 

[IRstuff]

An obvious possibility is if your modulus is 10x higher than what you modeled.

TTFN

FAQ731-376

 

[GregLocock]

Biggest problem is often the fixity of the constraints, especially encastre. These days I usually model the test rig as well as the test item.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[flash3780]

For most beams, Bernoulli-Euler beam theory is going to give you results that are very nearly dead-on. I would recheck your calculations, material properties, and your measurements.

As was mentioned, the assumptions for a cantilever beam don't work as well near the fixity, so for a very short beam, things can start to fall apart. Can you give the dimensions of your beam?

 

[hokie66]

Until and unless the OP provides more information as requested, I suggest we stop speculating about his nebulous problem.

 

[FeX32]

....period


Fe