Lab 2 Activity

Lab 2 consists of 3 main parts:

• Learn to use Simulation-Xpress within SolidWorks to evaluate structural properties of the parts you design.
• Verify the correct behavior of Simulation-Xpress on a simple beam loaded two different ways as compared to analytical expectations.
• Design a flexure part to satisfy certain conditions on flexure without breaking or being too flimsy.

We will design the flexure parts in lucite/acrylic/plexiglass so that we would in principle be able to easily fabricate designs via the MAE laser cutter facility or via a water-jet cutter at the Scripps shop.

The procedure in more detail

Note: see Tips section below for helpful hints.

1. Learn how to use Simulation-Xpress. Tutorials are built into SolidWorks, but the machines in the lab may not allow these to work. If you want to try, go to Help → SOLIDWORKS Turorials; and in the 3x3 matrix of options, pick the SOLIDWORKS Simulations Tutorial. Within this, select the Static Analysis of a part. Even if not following along successfully, it may be useful to page through the steps to get an overview of the process. The problem seems to relate to Add-Ins, so as a workaround, you can just use this custom tutorial page to suit our purposes.
2. Create a simple beam in SolidWorks made of aluminum 6061-T6 a common, strong alloy), assigning dimensions 500 mm long, 50 mm wide, and 2 mm thick. (You can deviate from these exact specifications if you desire, but this is roughly what we're looking for—one thing you may want to do is find some similar piece of metal and model that for real-life comparison.) Be sure to save the part, or Simulation-Xpress will fail.
   1. Use SolidWorks to determine the moment, I, which you will compare to your calculated value. This is done via Tools→Evaluate→Section Properties, after having selected the appropriate face (in this case the smallest end-face; note tip below about selecting other). The I_x and I_y (in mm⁴) are the ones you want (which one?...what does the other one mean?).
   2. Test the beam flexing under a uniform load: pick the small end-face as the "fixture" (making the beam a cantilever; "select other" again may be useful) and put the load on the "upper" (largest) surface (something in the ballpark of 2 N may be appropriate, or you can use the actual weight of the part to mimic self-induced deflection; can use Tools’Evaluate→Mass Properties to get). The force will by default spread uniformly across this surface.
      1. Does the part dip below a "sensible" safety factor of 2 anywhere? If so, where? (Note: the safety factor quantifies how much margin you have before exceeding yield stress. Higher numbers mean more safe. A safety margin of 2 is commonly used.)
      2. note the maximum stress, where this maximum stress is, and the maximum deflection (in mm).
      3. calculate the expected maximum stress and deflection for the aluminum part you designed (analytically, then convert to numerical value). These should agree with the Simulation-Xpress results. Note: SolidWorks uses E = 69 GPa and σ_y = 55.15 MPa for Al 6061 (275 MPa for Al 6061-T6).
   3. Test the beam flexing under an end-load, again picking the small end face as the restraint (cantilever). Pick the end surface as the load surface, and make sure the load points "down"—or perpendicular to the beam length. Start by applying the same load as before, but check the safety factor carefully and if dipping below 2 anywhere decrease the force (proportionally) until the safety factor is satisfied. This will ensure that the beam stays in the elastic regime. Perform the same analysis and comparisons as for the uniform-load case.
3. Pick one of the designs here, and design a part that satisfies the associated goals. The goal is typically to achieve a flex to the specified deflection amount, without exceeding the threshold stress anywhere, and simultaenously hitting a "safety stop" when the design deflection is reached. First sketch out the part and perform analytic calculations to guide you into the right ballpark. You'll find this is much more efficient than letting Simulation-Xpress tell you when you're getting close. (The analytic calculations should be represented in the write-up.)
   1. Assume the part will be cut out of a sheet of lucite (plexiglass/acrylic) 6.35 mm (0.25 inch) thick. The machining tolerance will be approximately 0.125 mm (0.005 inch). The path will be under computer control, so any weird dimensions will work. Make sure that the vane thickness (in the vertical direction) stays above 1.5 mm, but try to stick above 2 mm, if possible. It's possible that the design goals I set out demand thicknesses less than this, but I hope not. Also, keep intentional gaps to 1 mm or larger.
   2. Note the standard outer size we are using for all flexure designs of 17×11 cm.
   3. Wherever practical, add structures that will prevent the part from moving beyond its design range. Little rectangular bumps, or even the frame structure that many parts have can perform this limitation.
   4. Be sure to design in a hole for hanging a mass at the appropriate position to load the part as per the design.
   5. Generally speaking, the entire part (even the hole) can be drawn as a 2-d sketch (using line, rectangle, and circle tools), then extruded. Use add relations to establish co-linear lines, equal line lengths, etc., so that you don't have to dimension every single thing. When the sketch is fully black, you're set... If you have been smart about your relations, etc., changing a parameter like beam width will require the change of only one dimension, and the rest of the part will accommodate automatically.
   6. You may need to design special small surfaces onto which the load or loads are applied. If you have made a hole for hanging a mass, you can even select this surface for the load, making sure to establish the direction of the forces all in one direction, rather than normal to the surface (this was also done in the hook tutorial).
   7. Create a custom material that we'll call lucite. To do this, first assign Plastics:Acrylic to the part. Then edit the material (create/edit button), select <New Material Database>, classification: Plastics, material name: lucite. Go to physical properties and double-click the elastic modulus to make 2400 → 3000 (3×10⁹ Pa), then double-click the yield strength to make 206.8 → 10.0 (10⁷ Pa). (basis for numbers)
   8. Compute the expected load for accomplishing the target deflection. Note that when there are multiple beams moving the same way, each will require the computed force, so that the total load is the sum of all the individual loads needed.
   9. Once the part is designed completely, perform the Simulation-Xpress analysis. Make sure the max deflection is in line with the design goal. If it is too large, decrease the load until this is matched. Look for any instances/locations in which the target safety factor of 2 is exceeded. If it is (and the deflection matches the goal), then modify your model to (just) meet the safety factor requirement. Note that the load will change as well. Keep the deflection at target as you iterate. Any safety factor between 1.80 and 2.20 can be called good enough.

Tips

Here are some tips that may be useful:

• One minute spent reading and understanding each tip may save you ten minutes or more in execution of the lab. What a bargain!
• It seems we must disable the Simulation Add-In in order for materials properties to work, which is crucial for simulation to work. To do this, go to Tools→Add Ins, and uncheck the Simulation entry.
• Use SI units for this entire lab. Select mm (rather than m) as the base unit (Options (settings gear)→Document Properties⤒Units→MMGS).
• It may be necessary for you to clear some junk out of your directory if Simulation-Xpress doesn't work for you initially.
• When starting Simulation-Xpress, make a habit of clicking Options on the welcome screen so you can place the results in your own folder (recommend making folder within Desktop to stash results), and also check the (useful) "show annotations for max/min" box.
• A useful trick for selecting a surface that is hidden from view is to right-click on the region, pick "select other" then select the desired surface. This lets you access faces behind the visible ones.
• The cartoon graphic of the flexure is way exaggerated, and in fact will look the same indepentent of actual deflection. Don't get tripped up over this.
• You may want to change the resolution of the analysis to verify convergence and the most accurate numbers possible (at expense of run-time). This can be done via Change Settings in the Simulation-Xpress wizard/guide. A sensible approach would be to make any design adjustments based on low-res results, then run the high-res once the design has stabilized (otherwise spend unnecessary time waiting for iterations to complete analysis).
• If including radii for strain relief, explore a bit how much radius you actually need using the Simulation-Xpress analysis. Making a fillet too large shortens the span of the beam, in turn demanding more from it. Often, a 1 millimeter radius will completely do the job.
• You can always make a beam thinner and thus avoid overstressing the beam, but the beam becomes significantly flimsier, and we want to keep it as stiff as possible so we don't worry about accidental breakage, and also can test in real life with reasonable loads. If the load required to reach maximum deflection is very much less than 1 N, its own weight will start to dominate.

Unsolicited Advice about Trusting Numbers

I tend to trust analytic calculations more than numeric results, and want to see that the computational tool is being used correctly: that it produces expected results. Students often trust the computer more (under the assumption that it knows more), and feel that part 2 of the lab is primarily a validation of your calculations—that it is more likely correct than you are. But computers will always spit out a number, even if garbage. I've been fooled too many times by a professional-seeming package—I need an independent check I can trust. If I don't know how to calculate something, I'll try to model a system that I can test physically, to make sure I'm on the right track and not wasting my time and misplacing my trust. And it's not always a matter of distrust in the product, as much as uncertainty that I know how to use it correctly. For good reason, such things are called "sanity" checks.

Write-Up

The write-up will consist of the following things:

• A description of the simple beam used for Simulation-Xpress verification (dimensions, moment I, etc.).
• The results of the analytical simple beam calculations, and comparison to the SolidWorks Simulation-Xpress results.
• The analytic calculations used to guide the design of the flexure piece (before the SolidWorks implementation).
• A print-out of the 3-d model of the flexure part in some favorable viewing angle to show it off well.
• The Simulation-Xpress results from the final analysis of the flexure part, including specification of loads, final dimensions of the flex parts, maximum stress (and identify location), and safety factor achieved for the part under load.
• A note about how the final design dimensions differ from the initial calculations (and why that might be so).
• Optionally, any reflections on the hard parts, the fun parts, etc.