A Case Study in Analysis: Verifying Strain Gauge Testing with SW Simulation

1
A Case Study in AnalysisVerifying Strain Gauge
Testing with SW Simulation

• Richard Wand, President
• MJ Engineering Consulting Inc.

2
A Case Study in Analysis

• What we were given by the customer
• Known Field Failures
• SolidWorks Models with Proposed Field Fixes

3
A Case Study in Analysis

• What we were given by the customer
• FEA Results done in-house of Original Design and
  Proposed Change showing significant improvement
• Good, well documented test procedure
• 9 different tests with original design

4
A Case Study in Analysis

• What we were given by the customer
• FEA Results done in-house of Original Design and
  Proposed Change showing significant improvement
• Strain Gauge Test Results of Original Design
• Strain Gauge Test Results of Proposed Change

Test 7 showed worst stress at gauge 3
5
A Case Study in Analysis

• What were we able to give the customer back
  (Jumping Ahead a Bit)
• Reasons why the in-house FEA was incorrect
• Revised design (with FEA improvement)
• Confidence to test revised design

• FEA 62 improvement (strain)
• Strain Gauge 44 improvement (max stress)
• Strain Gauge 55 improvement (stress range)

6
A Case Study in Analysis

• What were we able to give the customer back
• Worst load from initial strain gauge testing,
  Revised Design showed 64 improvement

7
A Case Study in Analysis

• A quick word about strain gauge testing
• Make sure your tests are well defined, consistent
  and processes are repeatable and well documented.
• Remove as many variables as possible, use loads
  and situations that are easily replicated in
  Simulation software.
• Understand the type of gauges used, the direction
  (if applicable), the data acquisition software,
  the sampling rates etc
• Prior to adding loads, run a baseline test, then
  run load tests and repeat baseline if baseline
  tests arent consistent and reliable, neither are
  results
• Understand the possible effects of weather,
  moisture, wind and other aspects of testing that
  can lead to uncertainty.
• Take pictures of the tests and processes for
  documentation purposes

8
A Case Study in Analysis

• So how did we get to the results presented??
• Started by Reviewing Customers SW Models
  Analysis Results (Paper Copies)
• Large sheet metal parts meshed as solid elements
  not shells (mixed mesh)
• Displacement plots showed inconsistent results
  between bonded members
• Load transfers (physics) didnt appear correct
• Strain Gauges compared to FEA results by plot
  legend/colors only

9
A Case Study in Analysis

• When to use mixed meshes?
• Typically Geometry Driven
• Beams Length compared to cross section, 20
  times or more
• Shells Rule 1 Surface Area to Wall Thickness of
  101 or greater
• Shells Rule 2 If the part would be
  understandable when modeled with zero thickness
  surfaces to someone unfamiliar with its actual
  form, a shell model is a likely candidate
  (Source Building Better Products with Finite
  Element Analysis by Adams and Askenazi)
• Sometimes it is Load Driven
• Shells (particularly thin shells) neglect normal
  stress in the thickness direction, so if bending
  is a major component of the load carrying
  capability, then thick shells or solids may be
  required
• When Thin vs. Thick Shells
• Use Thin shells when thickness to span is less
  than 0.05
• Shells can be a pain use selectively!

10
A Case Study in Analysis

• Why use mixed meshes?
• Accuracy, stress and displacement values can be
  off when using solid elements on geometry more
  suitable for shells.
• Takes more solid elements to accurately model the
  same results that can be obtained by shells.
• Speed, Speed, Speed and more Speed
• Shells generally run 10-20 times quicker than
  solid elements from my experience
• Results can be easier to interpret and visualize,
  but sometimes harder to explain to customers!

11
A Case Study in Analysis

• So HOW do we use Simulation to verify design
  into Solidworks we go.
• Shells, how to create and some tips to remember
  thickness
• Split Lines
• Contacts, Global and Local
• Mesh Controls
• Restraints on Shells, particularly symmetry
• Tips for running mixed meshes (soft spring
  options)
• Sensors
• Checking physics by looking at displacement plots

12
A Case Study in Analysis

• Quick Look at the Original FEA Models

13
A Case Study in Analysis

• Shells, how to create and some tips to remember
  thickness

14
A Case Study in Analysis

• Split Lines

15
A Case Study in Analysis

• Contacts, Global and Local

16
A Case Study in Analysis

• Mesh Controls

17
A Case Study in Analysis

• Restraints on Shells, particularly symmetry

18
A Case Study in Analysis

• Tips for running mixed meshes (soft spring
  options)

19
A Case Study in Analysis

• Sensors

20
A Case Study in Analysis

• Checking Physics by looking at displacement plots