Stress Concentration problem in Ansys Workbench 15.0 2

[Bilal91]

Hey,

I got a problem with stress concentration after I applied a point force.
The tool has a fixed support at the half of the shank approximately.
Attached is a pic about the problem.
How to approach to fix that problem.

 

[corus]

If you apply a load at a point, that has in effect zero area, then the stress is infinite. Try applying the load over a larger area.

 

[rb1957]

this is what you get when you load a point. in your case it looks like you can't rationalise an area, so a material non-linear (elastic-plastic) analysis is required ... if you really need to know the stress on the edge of a cutting tool ...

another day in paradise, or is paradise one day closer ?

 

[Bilal91]

I tried to increase the area but unfortunately it didn't work.
By the way, I am using Tungsten carbide cobalt (10%) as a material.

 

[rb1957]

is this a linear analysis ? (i think so) then you're chasing a phantom (linear models extrapolate stress linearly beyond yield, and so produce fictious stresses.

how'd we solve this before we had models ? we'd make drill bits, and see how they worked, and learnt which designs worked and which didn't.

imo the best way forward is to model an existing drill bit in the material you're using to see what stress has been shown to be acceptable, then judge your stresses against this standard.

another day in paradise, or is paradise one day closer ?

 

[Bilal91]

Sorry, I can't get you exactly what you mean.
Can I use a Moment instead of a force ?
Would it solve the problem and give good results ?

 

[rb1957]

moment ? who said anything about a moment ?? what i said was in the past (before models) we learnt by doing (seeing waht worked). now you can reverse engineer this ... model something that works and the stresses produced are reasonably acceptable for the new design.

but it does raise another question about your analysis ... where is the load coming from ? how calculated ??

another day in paradise, or is paradise one day closer ?

 

[Bilal91]

My model is so easy. Isotropic Elasticity model using tungsten carbide cobalt as a material.
Applying a radial force near the tip of the tool while fixing the clamp.

My question now, can I use a Moment instead of the force ?

 

[corus]

To apply a moment you'd have to apply two point forces, so your problem doesn't go away. For the radial load you've applied you'll get a bending moment at the clamp. Check if the bending stresses are what you'd expect at the clamp simply by adjusting the contour values. At the tip you'd expect to see ridiculously high stresses because of your point load, but elsewhere they should be reasonable - if you've applied the correct load.

 

[Tmoose]

Is the "point load" derived from the known power required for machining?
Or was the load given in the problem description?

===========

Even if this is an exercise I'd be wondering what the material being cut thinks of being asked to provide this "point load". Like, how can I (the workpiece) generate a point load capable of inducing a principal stress over 80,000 psi.
Big loads suggest we may be talking about a heavy roughing pass, and that generally means "increased tool engagement angle" with the likelihood of significant length of tool edge engagement and even multiple tool edges.

http://www.conicalendmills.com/wp-content/uploads/2014/10/Machining_Methods.pdf

Cutting tools generally show wear beyond a point. More like the left hand picture here.

http://lcamtuf.coredump.cx/gcnc/07-damage.jpg

 

[rb1957]

"Isotropic Elasticity model" = linear elastic ?

why would you think to apply a moment (if you're intention is to apply a linear force) ? if this is a type of cutter, the cutting force is off-axis and so there is a moment induced in the part, but i don't see a moment being applied, not even a couple.

if you're analyzing a part, how do you know that 80ksi is no good ? even if it is above yield (or ultimate) it is not a real stress (if you're using a linear FEM). if it is higher than similar parts in service then maybe you've got a reason to question it.

another day in paradise, or is paradise one day closer ?

 

[Bilal91]

The thing is, I don't know the value of the force. That's why I am using FEM to know what maximum force I can apply on the tool before breakage.

I tried to apply a force on a bigger area (many nodes together) as attached below.

Are the results logical ?

 

[rb1957]

if you're looking for part failure, then you need non-linear FEA.

if you're looking for part failure, then the stress peak at the load application "should" be irrelevant, unless you're looking for the tip of the bit to clip off.

for a solid, I'd use von mises rather than max principal.

is the scale showing that the peak stress is somewhere else ?

another day in paradise, or is paradise one day closer ?

 

[inline6]

doubt FEA is the right tool for the job

 

[Bilal91]

I want only to check where failure may occur without seeing the failure behavior and how the crack propagates.
Only analysis through the contour colors.

Besides, I have the Transverse Rapture Strength (TRS) value (3700MPa), Poisons ration, elastic modulus and K1C only.

Can I compare TRS with von mieses to find the critical areas ??

 

[rb1957]

"Can I compare TRS with von mieses to find the critical areas ??" ... i think so ... von mises combines together the three pricipal stresses of a body into one stress. if the "transverse rupture strength" is significantly less than ftu then using TRS would be conservative; more accurate would be to use ratios like in-plane principal stress/ftu and out-of-plane principal stress/TRS.

i still think you're chasing phantoms if you're using linear FEA.

another day in paradise, or is paradise one day closer ?

 

[Bilal91]

What is the difference between Von Mises and maximum Principal stress ?

I have a Brittle material (WC-10% Co), and I would like to see when it fractures. I have Transverse Rupture Strength value (3700 MPa).

Can I compare the Max Priciple Stress with TRS with ignoring Von Mises ??