Notes about
the ANSYS simulation of the loaded pedal crank:
Because the purpose of this assignment is for you to interpret your ANSYS
results in the context of your knowledge of strength of materials and boundary
conditions, this section will not discuss those issues. However, you
may want to make note of the following issues regarding the ANSYS model you
have extracted results from.
1) The dimensions and shape of the pedal crank were drawn in Pro/ENGINEER
computer-aided design software and then imported into ANSYS finite element
software for the numerical modeling. The drawing of the part (done
here at CMU) approximates the shape of an actual pedal crank by matching
the most critical dimensions, but does not duplicate all of the actual dimensions.
2) In the pedal crank model, two concentrated loads are applied at
the pedal end of the crank, with a net downward force of 200 lb and a net
moment of 200.0 lb x 2.219 inches = 443.8 lb-inches (equaling the net force
and bending moment applied by a 200 lb rider with his full weight on one
pedal). The pedal crank has all 3 displacements constrained to equal
zero at all points inside the hole on the crankshaft end (roughly approximating
constraint from the crankshaft, which fits into this hole). St. Venant's
principle says that it is ok to apply these "statically equivalent" loads
if we are only interested in results away from where these loads are applied.
3) The ANSYS numerical model is approximating all of the stresses that
occur in a pedal crank that is loaded as described above. In other
words, it does not assume beam or torsion theory in getting its predictions.
In this way, you could say that the ANSYS model is providing a more complete
picture of the actual stress state in a pedal crank (read below, however).
4) The numerical model should be giving reasonably accurate results
for stresses away from sharp stress concentrators (i.e. away from sharp transitions
in the geometry such as near the pedal crank ends). The accuracy of
the ANSYS model in regions of sharp changes in geometry is suspect because
such regions require a high density of elements and nodes to obtain accurate
results. Multiple models having various element/node densities in these
regions were not run for this problem to ensure that the density used was
sufficient to obtain fully accurate results. Thus, don’t be convinced
that the sophisticated numerical model is “correct” near the sharp changes
in geometry at the ends of the pedal crank just because it comes from a computer
and models the full 3-D problem.
Note also that the loadings applied to the pedal end of
the model are idealized “statically equivalent” force and moment loadings.
Also, the full constraint condition applied to the crankshaft end of the
model is an approximation of what really occurs at that location. As
a result, stresses near the ends of the pedal crank model will not be realistic
regardless of the element/node density used there (because the loading and
constraints at these locations is not realistic).
Given these facts, results from the ANSYS model near either
end of the pedal crank are not accurate.