L shape beam

[Ignicolist]

Hello,

Im trying to design a casted component with the following shape.
As expected, I find high stress on the corner and a radius can reduce them.

My question is: If the maximun stress on the radius is higer than yield strengt,
in static: It will "plastify" a little and this high stress will be redistributed
in dynamic: I will go to the jai after killing people because this componente will broke.

In addition, I would like to know if there are more posibilities or solutions. As the stress with the round was too hing, I tried with an undercat (asuming that the stress will improve but the deformation will be bigger.
I also investigate the posibility of other shapes (like two differend round, or parable shape) but nothing seem to work.

Thanks in advance!

 

[corus]

It'd be better of you provided a drawing of the shape and if possible pictures showing the stress distribution. In general, however, a stress concentration would be classed as a peak stress and should be assessed against fatigue damage when there are varying loads, using an appropriate design code.

 

[Ignicolist]

Here you can find some pictures.
This part is fixed and there is a force applied on the marked disc. And the results are "as expected", but not good enough...

I can´t increase the round on the critical point because there is a "forbidden area" which has to be free of material

 

[kingnero]

enlarge your radius by moving eihter one of the planes?

 

[Ignicolist]

Thanks for the answer kingnero.

The "Forbidden area" (FA) is fixed. Then, if I move the plane under FA, the component will be thinner and its inertial will be smaller.
If I move the other plane, the vertical one, I will increase the lever arm.

Moving one (or both) planes nearly the same as using a an undercut.

https://www.google.es/search?q=din+...%2Fmanual.kisssoft.ag%2Fen%2F8251.htm;400;271

I´ve already tried with them and the improvement is small or nothing, as the increased radius has to deal with the above mentioned problems.

 

[dwallace1971]

Is the load you're subjecting this to a limit or an ultimate load?

"On the human scale, the laws of Newtonian Physics are non-negotiable"

 

[Ignicolist]

THe load that I use (130kN) is the maximun load on the application. Then, I consider the limit of this material (GJS-400-18) and a safety factor.
Notmally, this application will work under 130kN, or without loads. Following cycles.

 

[dhengr]

Ignicolist:
If you increased the height of the red boss on which the loads are applied or the attachment is made, you could increase the radius back at the joint btwn. the horiz. and vert. legs of the angle bracket, without changing much else. You might also shape the part which applies the vert. load on the boss, on its side near the vert. leg of the angle bracket, so as to tolerate a larger radius on the joint btwn. the horiz. and vert. legs. Running the radius further up the vert. leg would probably help a bit, an elliptical shape. The stresses on the surface of the radius are as you would expect to see them, max. in the middle and decreasing as you move away toward the sides; and you almost can’t make the radius large enough to tame them. If you really have a fatigue issue, I would shape and polish this radius surface. And, the radius surface at the sides of the angle bracket may as likely be the start point of a failure crack, if all stress raisers aren’t removed. Since you are casting this, you could gradually increase the radius as you moved away from the middle, and this increased stiffness would draw some of the moment stresses away from the area (middle) which shows red now, and average the stresses out a bit. That high stress region is really fairly will confined, so a bit of yielding shouldn’t be a killer unless you also have a high cycle condition. The fatigue problem needs more detail, study and testing. FEA almost always show a very high stress in that region, plays a bit of a trick on you, even if the stress level is tolerable. That region, of that detail, is always difficult to rationalize.

 

[hendersdc]

Ignicolist,

Would you be able to add gussets to the sides, out of the way of the applied load?

 

[corus]

Perhaps it depends on what limit you are applying to this stress. The stress at the fillet is due to a stress concentration and as such limits based upon yield stress wouldn't be applicable as the failure criteria for a stress concentration would be due to fatigue damage. Mean stresses through the section, and the linearised bending stress limit would be related to yield, however. For this stress classification the limit would depend upon the number of cycles the part was subjected to, ie. it's required fatigue life. In the extreme the limit could be for infinite life and the limit is generally about half the UTS figure. You'd need to check if there is any data on the fatigue life of this material to check on the true figure.