Brittle Material Stress Analysis

[gkemner]

Has anyone used Mechanical to analyze the stress in a brittle material? Such as glass, crystal, ceramic or something similar. Any problems or special considerations that I need to take into account to get accurate results?

Thanks!

[randyspeed]

Greg,

There is no place to handle a Brittleness property in MECHANICA, other than
to enter an allowable stress. Just be aware of the fatigue stress limits
(S/N curve) of the material you are analyzing.

Randy Speed
President and CEO
Speed Consulting, LLC - "MECHANICA Experts"
(972) 938-0490 ph (972) 937-2319 fax
www.speedconsulting.com

[jholst-2]

Make sure you do not have any non-tangent inside corners in your model.
Include all fillets even if they are very small. You may want to add some
mesh control in the fillets. In real life, brittle materials will fail in
these areas and are not as forgiving as ductile material. Stresses
calculated in these area should not be ignored. Other than that there is
not anything special about FEA and brittle materials.

---Jim Holst

[ptc-442673]

Greg,

some additional points:

* use for evaluation max. principal stresses or Tresca (2 times max.
principal shear stresses)
* crystals and ceramics often have anisotropic behavior; use
orthotropic material definition (and orientation)

Be aware of the big scattering of the max. allowable stresses of such
materials (e.g. if you use Si or GaAs the average allowable stress is
e.g. 100 MPa, but to get a - bad - 10 % crack probability you should
allow only about 10 to 20 MPa !).

With best regards

Stefan Reul




Kemner, Greg (Space Technology) schrieb:

>Has anyone used Mechanical to analyze the stress in a brittle material? Such as glass, crystal, ceramic or something similar. Any problems or special considerations that I need to take into account to get accurate results?
>
>Thanks!
>

[david.reid]

Surely it necessary to consider vibration?

After all brittle materials will mostly break under impact or impulse load.
Can be very strong otherwise.

I can think of a case when someone did a stunt that involved balancing a
steam roller on 4 pint glasses, and they survived.





---
David Reid
Optima Design Services Ltd
davidr@optima-design.co.uk

[randyspeed]

I was thinking of that same thing this morning - I believe you are right
David. Fatigue is less important than just keeping stress down to an
absolute minimum.

Someone mentioned it earlier and was right on (didn't keep the emails) -
sharp (reentrant) corners are extremely bad. Must maintain very generous
fillets. Reminds me of a glass top I had made for my desk. I initially had a
rectangular penetration for cables and such (very small fillets). After
replacing it two times, each one getting larger fillets, I eventually just
made a half moon slot on the edge of the part.

Randy Speed
President and CEO
Speed Consulting, LLC - "MECHANICA Experts"
(972) 938-0490 ph (972) 937-2319 fax
www.speedconsulting.com

[r1chardgreen]

I've read the views expressed on brittle materials with interest and,
with apologies for stating the obvious, would like to add the
following to the discussion:

- Brittleness in a material is effectively the absence of the ability
of deform plastically.
- Brittle materials, when subject to tensile stress, will generally
fracture at the elastic limit or after a very small amount of strain
has developed beyond this point.
- Brittle materials are often much stronger in compression than in
tension (four pint glasses and a steam roller !! I would like to
have seen that David).
- The maximum principal stress theory is the most appropriate
criterion to use for predicting failure. (This explains the need to
avoid sharp corners Randy).
- The wide range of allowable stresses quoted for brittle materials
can be directly related to the "quality" of the material. (Type
"glass allowable stress" into Google and you'll see what I mean !!)

It is clear that great care should be taken to seek out all potential
stress raisers in the component under consideration and to use
appropriate mesh refinement to ensure that stresses in these areas
are predicted as accurately as possible. It follows that
"defeaturing" should be avoided at all costs!! The selection of an
allowable stress which takes account of the specific grade or class
of material used is clearly also very important.

Richard Green
Technical Director
Finglow Consultants Ltd.
consult@finglow.co.uk
Tel: +44(0)1992 550 700
Fax: +44(0)1992 550 843

[doug.bittner]

Hi All,

Might I suggest, rather than maximum principle stress theory, you might
consider modified Mohr theory, or perhaps Coulomb-Mohr theory. See most
any Strength of Materials textbook. Joseph E. Shigley gives a good
explanation in his text, Mechanical Engineering Design (page 254 in the
Fifth Edition), McGraw-Hill, 1989.

The Failure Index, of the Mechanica Results Display, will handle the
modified Mohr theory. Just be sure to enter the correct failure
criteria in the material properties set.

Yours,
Doug


Doug Bittner
Staff Mechanical Engineer
Danaher Precision Systems
Danaher Motion
200 Flanders Road
Westborough, MA 01581-0200