Stress At Hole in Plate

[Matador]

We have had several failures in our bale press and after some study the loading can be simplified as follows:

Visualize a rectangular plate with a hole in the center. The plate is held down in the center by one nut (super bolt) in our case. There is uniform loading (not necessarily equal on each side) along two opposite sides only. We experience numerous plate failures at the hole where the plate material is at a minimum.

I believe it's a fatigue problem because the load changes every time the press cycles. Is this strictly a bending stress anaylsis (Mc/I) or do I have to use some form of thin membrane theory? The plate in question is two inches thick and there is about 25 ton force on each side of the plate.

This cry for help is from a Mech eng, stuctrually challenged.


Thanks

 

[zcp]

A metallurgist could tell you exactly what the failure mode was and probably give you some suggestions to rectify the problem (new material, different supports, etc.)

And / or you could obtain some consulting help to determine your issue and set a course for solving the problem.

ZCP

http://www.phoenix-engineer.com

 

[boo1]

What is the dimensions of the plate and hole, whats the plate material, whats the finish on the hole, are the edges champfered, cycles at failure, and loading sequence? Can you prove a picture?

Sounds like you have bending of a finite width plate with a single hole. The stress concentration factor Kt must be accounted for (See Peterson's Stress Concentration Factors second edition Chart 4.83) or peform a FEA model analyis

 

[Matador]

The 2" thick plate is 14" wide x 15" long with a 6-5/16" dia hole in the center. The uniform loading is along the 14" wide sides.

The nut for the superbolt in the center rests on a bearing plate (washer) which is approx 9" dia. The maintenance people have already upgraded the plate material once to T1 but I have not received the specification sheet yet from the manufacturer.

After the failure analysis, it will come down to design and material selection. Perhaps higher strength material is not always best for fatigue failures. I'm convinced the problem is the repeative loading of the bale press as the cylinders compress the material, then relase the load until the next bale comes along.

I can combine the uniform loading along the edges into one force and then calculate Ixx for the different cross sections and use stress=mc/I but this may not be that simple.

 

[rb1957]

i think part of your question is "is a thick plate, ie plate bending (Mc/I), assumption appropiate ?"

that would depend on the displacement under load. if the ends of the plate are deflecting less than 1/2 the thickness of the plate the plate bending is dominate, more than the thickness and you've got a reasonable amount of membrane laoding.

i'd suggest that a large deflection isn't what you want, so maybe stiffening the plate with some "angle iron" would help. as for your failure, like the first responent says some metalurgy analysis of the failure would confirm the presence of fatigue. And like the second poster, you'd have to include the Kt effect of the hole (i'd start with 3, but strictly this doesn't apply to your geometry). but you need a good nett area stress to start with.

some other questions ... has this press been working fine untill recently ? has something changed, that would initiate failures ? can you reinforce the hole, either with angles, or a doubler ?

good luck !

 

[miecz]

How is the hole made in the plate?

 

[Matador]

I'd say the plate deflection along the edge where the load is applied is ~ 1/8".

The plate in question fails about every 9-12 months. There is a lot of potential for damage when one of these plates fails beacuase the load has to be applied with 3 cylinders instead of 4.

 

[zcp]

How many loading cycles in the 9-12 months?

ZCP

http://www.phoenix-engineer.com

 

[boo1]

Still need:
"..whats the plate material, whats the finish on the hole, are the edges champfered, cycles at failure, and loading sequence?"

Kt=2

 

[rb1957]

sounds like plate bending is a reasonable assumption ...

question, it sounds like you've got one of these guys on each corner of the bale; why apply the load on two sides only ? (just curious) ... i'd have thought that the entire surface of the plate was bearing up on the bale.

if this has been happening since the year dot, it does sound like a design problem. btw, how thick is the bearing plate (that is acting like a washer under the nut); i'd have thought it should be thicker than the plate (>2").

when this goes "tits up" (breaks) is the component (i'm guessing its this plate we're talking about) expensive ? if not, why not replace every 6 months (to prevent in-service failures). if there're four on each press, does a particular one fail, or randomly distributed between the different corners?

The stress at the edge of the hole is not trivial to calculate, at least not without some thought. waht M to use ? you know the force in the press, this is distributed by the washer into the plate. as this is some distance from the edge of the hole (washer diameter 9", hole diameter 6"), maybe not much is happening at the edge of the hole ?? it's not like the classical hole in a tension loaded plate, where the tension stresses have to go around the hole.

back to your original post, you're right that high strength steels at reduced fatigue properties, or at least they are much more susceptable to crack initiation from inherent flaws, but i suspect that your steel is pretty
"agricultural", maybe 125 ksi ... the problems happen with the ultra high strength steels, 240 - 300 ksi.

 

[Matador]

The press cycles about 2300 times per day. Each bale is ~ 484 lbs & we produce ~ 550 ton per day.

There is one hugh cylinder in the center which starts the initial compression. Once the initial compression is complete the other four cylinders assist to apply the final force to compress the bale to the desired size.

The nut in question is called a superbolt nut and it comes with a bearing plate. Basically the nut screws on to the large diameter (~5") bolt. Then there are individual small screws (~8) all around the circumfrence of the nut which get tightened down on the bearing plate supplied. This design allows us to apply a lot more tightening torque than could be achieved with a standard nut.

I don't want to do too much to the surface where the bearing plate rests on the plate but perhaps a bit of radius would not remove too much of the bearing area.

 

[jen4950]

I'd imagine you're using A36 steel (2" plate). 36ksi yield, 50ksi tensile.

Can you go to a thicker plate? Five inch diameter bolt intuitively would be a whopper for 2 inch plate. As a point of reference, we did a building in downtown Dallas in the 70's- it has 6 x 5 inch diameter anchor bolts (on a braced frame bay, columns are 1100 pounds per foot), and the A36 plate was specified as 10.5 inches thick. Of course the plate was 65" square and had a massive column on it, but a 5 inch bolt is not something you find at the hardware store.

So you are looking at about 500,000 cycles- thinking back to my grad school days, it's a function of change in service load stress that determines the cycles until failure. So if you are getting early failure due to fatigue, that fundametally means you are getting too close to the yield stress of the base material on each cycle- only way to fix that without reducing the load is to add more material to resist the load.

 

[corus]

Clearly the problem is due to fatigue as failure occurs after a number of cycles. The stress in the plate doesn't have to be close to yield to cause fatigue failure but depends on the stress range and the mean stress. The endurance limit of a material is related to the UTS of the material so a change of material would help. If the plate is bending then increase the MoI to reduce the stress or consider additional restraint, ie more bolts.

corus