[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PSUBS-MAILIST] Hemisphere Seat Formula



The seat doesn't care what shape the window is, other than the geometry of the contact faces.  The only effect that window shape has is how the window deflects under pressure, which places the stress on the inside diameter in the case of flat viewports, and increases the radial forces in the case of spherical ones.  Doing a very simple analysis off the cuff, In the Stachiw page 526 example, I'm looking at a projected diameter of 6.932 inches, giving a surface area of 37.740 in^2.  At 4000 psi, that is 150 962 pounds force.  The seat diameter at 6.932 inches and with thickness 0.125 inches, gives a total shear area of 2.722 in^2, so your total shear stress is 150 962 / 2.722 = 55 460 psi.  Within the 80 000 psi yield, but not a huge safety factor.  In any case, this is just an estimate, and I'm sure that Stachiw knows what he is doing.  I didn't look at your calcs - I don't have sufficient time right now, but I'll have a look this evening to see what you did.

-Sean


On 14/06/2011 10:58 PM, Jon Wallace wrote:
Sean,

That sounds like the same formula you sent me 4/9/2010 (see below) when I was looking at flat disk viewports, but it doesn't appear to work correctly for sphere segments.  The results I get using that formula indicate dimensions that are much larger than what I've seen in the field.

sigma=(W*l)/Z
=(W)[(D_o-D_f)/2]/[(D_f*pi)(K2)/6]

I'm using values from Stachiw's book, page 526, in which he gives dimensions for a 120 degree sphere segment used in his testing.  It includes the viewport and the mounting which he tested to 4000psi (see page 527, fig 11.80).  The actual seat diameter was only .125 inches.  Using the following dimensions from the book and the formula above:

Do = 6.932
Df = 5.97
K = .125
h = R * (1 - cos(theta/2)) = 2 (assuming 8 inch sphere)
Surface Area (SA) = Do * (pi) * h = 43.55
P = 4000 psi

Numerator:
W = 4000 * 43.555 = 174220
Do - Df / 2 = (6.932 - 5.87) / 2 = .481
giving
174220 * .481 = 83799.82

Denominator:
Df * (pi) = 5.97 * (pi) = 18.755
K^2 = .125^2 = .0156
giving
18.755 * .0156 / 6 = .0488

Result:
83799.82 / .0488 = 1,717,192

In this case, Stachiw used 80ksi yield steel for the mount and the numerator alone is above that yield point.  I know Stachiw talks somewhere in the book about the base of sphere segments contracting inwards under pressure some small amount, but enough to make a chattering noise if the seat does not have a gasket to act as a buffer as the acrylic compresses.  He describes it as more than one pilot being startled thinking the acrylic had cracked.  In any event, that seemed to indicate that some amount of pressure on the segment is radial instead of axial (against the seat) and using the entire surface area might be very conservative, and in some cases, too conservative.

Jon


On 6/14/2011 3:28 PM, Sean T. Stevenson wrote:
Jon - this is exactly how it is done - calculate the projected area of the dome, find the force, divide by the seat bearing area to find the stress on the seat.  The only caveat is that the stress will not be evenly distributed, so to find your seat thickness you need to assume that the entire force acts on the inner diameter to get the worst-case scenario cantilever.

-Sean