Design of Elliptical Pad/Cover for sight glass

The Code has provisions for elliptical openings in shells (UG-36 & 37) and for flat covers of non-circular shape (UG-34(c)(3)). Bolting can be designed using the methods in Appendix 2-5. However the flange thickness is not covered by the Code and the provisions of U-2(g), good engineering practice, apply. One way to model the clamping flange is to consider the long 'side' of the elliptical flange as a fixed-fixed beam, with a length equal to the largest outside dimension, under uniform torsional load from the bolt / gasket reaction. Both shear stress and twisting deformation should be considered.

Hi, Lawrence!

Good answer, for which I intend to awarded you my GA after you add just a tiny bit more clarity for me.

First, how does one measure, draw, and form the ellipse with the opening for the glass attachment's base and top to fit the shell? I think this might have been part of the original question. For this, do you need to know if the shell is elliptically influenced, elongated or round in shape, and vertical or horizontal? Would you just form a mold of the shell's surface at the point of intersection and copy it when you make the reinforcing patch, for example?

We haven't been told where the elliptical component of the sight glass is. it could be vertically elliptical through its cross-section to proved an optical illusion or magnification; or for some strange reason, perhaps the shell's shape or piping in the way, elliptically horizontally along the length of the 'long' (or 'both' if it's a manometer) side.

Interesting that you included possible torsion deformation of the glass during installation/maintenance change-outs due to a bolt-gasket/sight glass binding action. I wouldn't have thought of that eventuality right off the top as you did; but although I recall designing them in class, I've only ever dealt with changeovers using pre-made components in the engine rooms of ships; and yes, they sure can break. In designing this unit from scratch, would you change the glass thickness/elasticity, breakage resistance because of its or its connections' shape or location?

Mark

Hi, Lawrence!

A few comments on the use of a flat pane sight glass through the shell as you have described its employment/installation.

"The glass face has to remain flat, but the inside surface could remain flat as shown or machined out to match the shell profile."

Good idea to get rid of the flat surface in a pressure vessel. But how would you arrange the machining process?

"the major axis of the ellipse would be oriented perpendicular to the longitudinal direction and the minor axis perpendicular to the circumferential direction."

This may or may not be useful depending upon the shape or orientation of the shell in use when considering a sight glass. Manways don't care about orientation. (Wish I could fit through a 20" X 12" opening. I was barely able to squeeze into a boiler firing chamber!)

If the type of sight glass that you have described is not useful in this application due to the orientation of the shell and the pressure dynamics of the opening's location, then perhaps the good ol' aluminum or brass tube glass might be recommended. So a GA for pointing out the pressure dynamics as a consideration of using the flat glass. .

Mark

Mark,

To get the cylindrical profile, you would have to go to a milling operation rather than a turning operation.

Lawrence

Hi, Lawrence!

Agreed. A copycat lathe would ruin the outside of the new piece by rounding it like the outside of the cutout. Put the cutout and new piece (minus all the glass and gaskets, of course) on a copycat milling machine and let the machine do the work. Main problem in milling is preparing the cutout to exact size for copying. So the first job would be to mill in reverse. Mill the cutout to match the new piece for perimeter dimensions. Then mill the new piece to match the newly milled cutout for the single-sided curvature.

However, it's important to point out to AVINASHUCE that you have answered the original question about considerations for using this type of glass based upon the horizontally and vertically useful placement on the particular application and the internal pressure considerations of each.

"Longitudinal pressure stress is exactly half that of circumferential stress, the major axis of the ellipse would be oriented perpendicular to the longitudinal direction and the minor axis perpendicular to the circumferential direction."

If using the elliptical glass will not show the application's internal information with any useful degree of tolerance because of the shell orientation (e.g. vertical shell, horizontal glass = short window of consideration) and the vessel is under any significant internal pressure then another visual indicator such as tube might be indicated.

An improvement on the model you illustrated might be the addition of a pressure-protected light bulb compartment under the window inside the shell to light up the meniscus on the gage.

Mark.

Dear Lawrence

you have provided the exactly correct answer,my application is similar what you have shown in the figure.

but some doubt still persist,how reinforcement requirement are met in this application,can you have any example,reference for such calculations.

looking forward for your extended cooperation

The ASME figure that pertains to sight glasses and other studding pad type nozzles is Fig. UG-40(a-2). In this figure, the thickness of the pad from the outside surface of the shell to the flange face is considered to be an integral reinforcing pad. There is no 'nozzle neck', tn = 0. I hope this helps.