Design Analysis
Push Button Casting P/N 29520 Analysis for Rees Inc.
|
Created By
Meadows Analysis & Design, LLC Checked By
Marc A. Meadows, P.E. |
Determine stress in part under 400 pound test load. Iterate geometry changes to achieve adequacy in design.
Analysis Type -
Static Stress with Linear Material Models
Units -
English (in) - (lbf, in, s, deg F, deg R, V, ohm, A, in*lbf)
Model location -
C:\Temp36\1636hbr
Rees provided the solid model, associated assembly files, and a material data sheet on Zamak No. 3. Our solution took the folowing steps: 1.) Using the original model, we loaded at 400#. The von Mises stresses exceeded the strength of the material. After plotting precision on this analysis, we decided to refine the mesh. 2.) In order to conserve resources, we utilized the symmetry of the part. We were able to get good precision at this level. 3.) Upon our suggestion, Rees modified the geometry to add cross sectional area. The part still had trouble with the 400 pound loading and especially in small radius positions directly above the spring window. 4.) We altered the part to increase these radii and apply a 200 pound test load. Results: The 400 pound loading given for testing this part is too much for the geometry to resist. On my suggestion, Rees altered the geometry, I added some fillets, and reduced the load to 200 pounds. As a result, I am confident that the casting will peform in service.
Static Stress with Linear Material Models may have multiple load cases. This allows a model to be analyzed with multiple loads while solving the equations a single time. The following is a list of load case multipliers that were analyzed with this model.
| Load Case | Pressure/Surface Forces | Acceleration/Gravity | Displaced Boundary | Thermal | Voltage |
|---|---|---|---|---|---|
| 1 | 1 | 0 | 0 | 0 | 0 |
| Default Nodal Temperature | 0 °F |
| Source of Nodal Temperature | None |
| Time step from Heat Transfer Analysis | Last |
| Type of Solver | Automatic |
| Disable Calculation and Output of Strains | No |
| Calculate Reaction Forces | Yes |
| Invoke Banded Solver | Yes |
| Avoid Bandwidth Minimization | No |
| Stop After Stiffness Calculations | No |
| Displacement Data in Output File | No |
| Stress Data in Output File | No |
| Equation Numbers Data in Output File | No |
| Element Input Data in Output File | No |
| Nodal Input Data in Output File | No |
| Centrifugal Load Data in Output File | No |
| Part ID | Part Name | Element Type | Material Name |
|---|---|---|---|
| 1 | Part 1 | Brick | [Customer Defined] (Part 1) |
| Element Type | Brick |
| Compatibility | Not Enforced |
| Integration Order | 2nd Order |
| Stress Free Reference Temperature | 0 °F |
| Material Model | Standard |
| Material Source | Not Applicable |
| Material Source File | |
| Date Last Updated | 2004/08/18-09:38:55 |
| Material Description | Customer defined material properties |
| Mass Density | 622e-6 lbf*s^2/in/in³ |
| Modulus of Elasticity | 10e7 lbf/in² |
| Poisson's Ratio | .03 |
| Shear Modulus of Elasticity | 48.54e6 lbf/in² |
| Thermal Coefficient of Expansion | 15.2e-6 1/°F |
| ID | Description | Part ID | Surface ID | Magnitude | Vx | Vy | Vz |
|---|---|---|---|---|---|---|---|
| 1 | 200/2 | 1 | 49 | -100 | 0 | 1 | 0 |
| 2 | 200/2 | 1 | 58 | -100 | 0 | 1 | 0 |
| ID | Description | Part ID | Surface ID | Tx | Ty | Tz | Rx | Ry | Rz |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Z-symmetry | 1 | 7 | No | No | Yes | Yes | Yes | No |
| 2 | Z-symmetry | 1 | 8 | No | No | Yes | Yes | Yes | No |
| 3 | Z-symmetry | 1 | 9 | No | No | Yes | Yes | Yes | No |
| 4 | Hole sliding | 1 | 50 | Yes | No | Yes | No | No | No |
| 5 | Hole sliding | 1 | 59 | Yes | No | Yes | No | No | No |
| 6 | Fixed | 1 | 40 | Yes | Yes | Yes | Yes | Yes | Yes |
Note that maximum precision is near 0.4. Target is below 0.1. This precision is a measure of the adequacy of the mesh to simulate the material. If this number is high, it means that there is an abrupt change across a given element resulting in questionable material behavior within that element. A low number would represent a mesh that was fine enought to capture the behavior. An optimum value is desired to save on computer resources and time resources.
While high stresses are present, they are very isolated. Let's look at how much of the part is seeing stress above the yield strength of 32 ksi.
Quite a bit of the legs are engaged in yielding. This is probably going to lead to some type of failure. We should check precision next.
Good precision. Maximum is extremely isolated. Bulk of model has great results.
Plot showing part as supplied and marking the areas of added radii. (Look for red ovals)
New geometry has added some cross sectional area. We have also put 0.01" radii in the spring window corners. Because we know that the part was destined for failure at a 400 pound test load, we have reduced it to 200 pounds. Note that the stresses are still high, but let's look closer.
We are only showing stresses above the 32 ksi threshold, and can see that they are very isolated. Let's turn on the rest of the part to look at that plot.
We want to illustrate how isolated the stress is in the corner. If a crack initiated in this area, propagation would be difficult because of the quick yield and surrounding low stressed areas. There would be no energy to support propagation.
Negligible, but interesting displacement values. These can often tell you if your loading and boundary conditions make sense.