PRESSURE PLATE ANALYSIS

1
PRESSURE PLATE ANALYSIS

• MECE 6362
• BY MOISES VASQUEZ
• 5/6/03

2
OUTLINE

• PROBLEM STATEMENT
• PROBLEM FORMULATION
• FEA MODELING
• FORCE FUNCTION
• SQUARE LOADING
• ALTERNATE CYLINDRICAL LOADING
• RESULTS

3
PROBLEM STATEMENT
The pressure plates, Fig.1, are used in water
regulating valves, Fig. 2. The plates may bend
during loading causing the bead to lift away from
the diaphragms producing a leak. The analysis
will focus on preventing excessive deflection and
yielding.
BOTTOM
TOP
BEAD
Figure 1
SCREWS
SQUARE HOUSING
PLATE
DIAPHRAGMS
VALVE BODY
Figure 2
4
PROBLEM FORMULATION

• APPLIED FORCE
• The total clamp force is provided by the 35
  in-lbs on the four 10-32UNF screws as follows
• D .19 in, Nominal Screw Diameter
• K .2, Nut Factor
• T35 in-lbs, Torque
• FLT/KD921 lbf, Clamp load of one screw
• FT4FL 3684 lbf, Total load of four screws

5
PROBLEM FORMULATION

• CONTACT AREA
• The square housing shown in Figure 2 contacts the
  top side of the plate in two places, total
  area.304 in2 as shown in Figure 3.
• The area was modeled in PRO/E by changing the
  accuracy to .0001 and then protruding a solid of
  .0002 inches. This thickness was made 250x
  thinner than the thickness of the plate to
  prevent interference with the FEA.

6
PROBLEM FORMULATION

• Plate Material
• The actual plate material is a carbon-nitrided
  SAE-1010 cold rolled steel. An example of the
  case-hardening is shown in Fig 4 and Table 1.
• Due to a lack of modeling the non-homogenous
  material, C1020 steel was used.

Figure 4
Table 1
7
FEA MODELING
VALVE BODY

• A standard stress analysis with a mesh size of .1
  was used.
• A collision of the plate with the valve body was
  used to simulate the contact Fig. 5. Coefficient
  of Restitution was set to Zero for both bodies.
  The valve body was constrained in all directions.
• The simulation frame rate was set at 2500/s to
  prevent penetration and increase accuracy of
  contact.
• The loading was accomplished with a step function
  as shown in Fig. 6 which was applied as a
  distributed load normal to the contact area.
• The simulation data was taken at .16 sec and at 1
  sec for the square loading to verify that the
  analysis was converging with respect to time.
• The rubber diaphragms were not included in the
  FEA.

PLATE
Figure 5
8
Force Function
Figure 6
9
Square w/Step Loading
Figure 3
LOAD
10
Square w/Step Loading
TOP
BOTTOM
MAX. VALUES
11
Square w/Step Loading
Deformation Scale5.4513
12
Square w/Step Loading
BOTTOM
TOP
13
Square w/Step Loading t1sec
TOP
BOTTOM
MAX. VALUES
14
Square w/Step Loading t1sec
DEFORMATION SCALE5.4513
15
Square w/Step Loading t1sec
TOP
BOTTOM
16
ALTERNATE CYLINDRICAL LOADING

• APPLIED FORCE
• FT3684 lbf, total load stays the same.
• CONTACT AREA
• The cylindrical housing shown in Figure 7
  contacts the top side of the plate in two places,
  total area .3454 in2 as shown in Figure 8.
• Loaded area is .0002 inches thick similar to the
  square loading area.

17
ALTERNATE CYLINDRICAL LOADING
Figure 7
18
Cylindrical w/Step Loading
.38 GAP
Identical to top
19
Cylindrical w/Step Loading
TOP
BOTTOM
MAX. VALUES
20
Cylindrical w/Step Loading
DEFORMATION SCALE 8.2575
21
Cylindrical w/Step Loading
TOP
BOTTOM
22
RESULTS
23
RESULTS

• The step function converges with respect to time
  at .16 sec. as compared with the 1 sec. FEA run
  on the square housing.
• The FEA converges at approximately Mesh Size .04
  inches.
• The cylindrical housing has 26.19 less Max. von
  Mises Stresses than the square housing.
• The cylindrical housing has 33.8 less Delta_Y
  Displacement (the amount the bead lifts off the
  valve body).
• The Minimum Factor of Safety of the cylindrical
  housing is 35.5 greater, but still yields C1020.

24
RESULTS

• The deformations of the bead occurred in the same
  area for both the square and cylindrical loading
  conditions.
• The actual cost of the cylindrical housing is
  almost 5x that of the square housing. This means
  that going to a cylindrical housing would require
  marketing approval.