Minimum Required Thickness of Cone at Small End

[acbdcl]

I need verify the effective area of reinforcement of a cone-to-cylinder junction at small end.
The Code address the calculation of the minimum required thickness of a cone in UG-33(f), but I cannot find any procedure to calculate the minimu required thickness of cone at small end to be used in equation 4 of Appendix 1-8.

 

[nickelkid]

A worked example is in:

Pressure Vessel Handbook - Eugene F. Megyesy

 

[SnTMan]

acbdcl, I haven't done these calculations in some time, but I recall that UG-33(f) is in fact used, with the diameter used being that of the small end of the cone. By finding the required thickness at the small end, the excess available for reinforcement can be evaluated.

See Appendix L-3.3 for a (mostly) worked out example.

Regards,

Mike

 

[acbdcl]

Unfortunately Appendix L-3.3 and Pressure Vessel Handbook - Eugene F. Megyes not address the calculation of the minimum required thickness at small end under external pressure.

Regards

acbdcl

 

[prex]

You have a single required thickness defined between the lines of support of a shell exposed to external pressure. So, if you have no intermediate stiffening rings in the cone, the required thickness is the same for the whole cone.

prex

http://www.xcalcs.com

: Online engineering calculations

http://www.megamag.it

: Magnetic brakes for fun rides

http://www.levitans.com

: Air bearing pads

 

[SnTMan]

prex, I can't agree. As I recall when evaluating the need for reinforcement the cone required thickness is calculated separately for both the large and small end, based on the local dimeter. The excess is then available for consideration as reinforcement.

As I recall the cone required thickness is calculated per UG-32(g) for internal pressure and UG-33(f) for external.

Appendixes 1-5 and 1-8 are invoked for each case.

If a cone is exposed to both internal and external pressure it of courese has to meet all the above.

The basis for oblique cones is given in UG-36(e) and (f).

Regards,

Mike

 

[prex]

SnTMan, if you follow the procedure in UG-33(f) for determining the allowable external pressure (and the required thickness) of conical sections, you'll see that it is entirely based on D[sub]L[/sub], the diameter at the large end. Things are of course different when dealing with internal pressure.

prex

http://www.xcalcs.com

: Online engineering calculations

http://www.megamag.it

: Magnetic brakes for fun rides

http://www.levitans.com

: Air bearing pads

 

[SnTMan]

prex, I think we have a bit of confusion here. I was speaking only of the evaluation of available reinforcement per Appendixes 1-5 and 1-8, which I took as the point of the OP.

My basis is the definition of "tr" in Appendix 1-8, which is only used at the small end.

Of course the cone must be at least as thick as the UG-32 and UG-33 calculations, based on the large end.

Regards,

Mike

 

[acbdcl]

The UG-33 addresses the calculation of critical pressure relative to buckling. The critical pressure is a function of the average diameter of cone.

The Appendix 1-8 addresses the calculation of reinforcement of a cone to cylinder juncture under external pressure.

ASME defines tr as the minimum required thickness at the junction, so UG-33 cannot be used to calculate tr.

I think that tr must be calculated as a hoop stress where the allowable stress is the allowable compressive stress given in UG-23.

tr = Pe*Ro/Sc

Sc is the value of B from the external pressure chart read with A = 0.125/(Ro/t)

I don’t have much experience with pressure vessels, so correct me if the assumptions above are not true.

Best regards

acbdcl

 

[prex]

IMHO you are extrapolating the code and this is not permissible.
I have to insist: if you have no intermediate stiffeners, there is only a single value for t[sub]r[/sub] for the whole cone as defined in UG-33. If there were intermediate stiffeners, then you would get a different t[sub]r[/sub] for every section between the lines of support, so different t[sub]r[/sub]'s at the large and small ends.

prex

http://www.xcalcs.com

: Online engineering calculations

http://www.megamag.it

: Magnetic brakes for fun rides

http://www.levitans.com

: Air bearing pads

 

[SnTMan]

prex, I don't have time to go through this in a systematic way right now, but my opinion is based on past programming work with an in-house program, and getting good agreement with Codecalc results. Also, see Appendix L, L-3.3.

acbdcl, I don't think a hoop stress approach is usable for external pressure.

Regards,

Mike

 

[prex]

SnTMan, L-3 does not indicate how to determine t[sub]r[/sub], and this is the only point of this thread. To actively contribute, you should provide a method for determining t[sub]r[/sub] at the small end of a cone under external pressure. I don't find it in the code.

prex

http://www.xcalcs.com

: Online engineering calculations

http://www.megamag.it

: Magnetic brakes for fun rides

http://www.levitans.com

: Air bearing pads

 

[jamesl]

I agree with Prex on this one. UG-33(f) is based on large end diameter to determine L/D and D/t to calculate tr. You cannot use UG-33(f) to calculate a specific tr at small end only.

acbdcl is mixing UG-23 and UG-28 or UG-33. UG-23 is for longitudinal compressive stress (not hoop stress) caused or axial load or lateral bending. UG 28 and UG-33 are for shell or cone under external pressure. The mechanisms are different. Shell and cone under external pressure are related to an unsupported length, while UG-23 is not.

 

[InFlames]

Hi to everyone!

I'm almost new designing with ASME VIII but If you pick up Pressure Vessel Design Manual you will find a procedure ( 4-2) in wich is evaluated your problem. The required thickness either for internal or external pressure is the required thickness of a shell under internal or external pressure with a diameter of "small end"

Almost all programs use a kind of virtual straight flange to evaluate this when cone is directly attached to a flange or a nozzle.

Maybe is not correct at all but it will work.

Hope my comment had been useful

"Non-flammable" is not a challenge

 

[SnTMan]

OK, back in front of a computer now, got in a few days vacation.

prex, my previous posts explain my opinion that 1) UG-33(f) can be used to calculate the required thickness at the cone small end for Apx 1-8 calculations, and 2) that I came to this opinion while A) updating an in-house program and comparing the results to commercial software and B) considering the language of the Code and the Appendix L example.

I might add that the "local" diameter is also used for calculating required and excess thickness when evaluating reinforcement of openings in cones.

It seems to me that if the Code gives a means to do a calculation, it makes sense to use it rather than to go inventing one, or looking elsewhere.

Now, I understand that we don't agree. That's fine, we don't agree.

Regards,

Mike

 

[acbdcl]

SnTMan, I read the code again and you are right about the maximum allowable compressive stress.

I read the UG-33(f) and based on the post on InFlames, I developed a small program (a Excel spreadsheet) to calculate the conical straight flange dimensions.

I check the results against the program COMPRESS Build 6162 and the results are fine for angles between 14º to 60º.

I test the results for SA-516 70 and SA-240 316L at 100ºF and 500ºF.

I'm translating this code from portuguese to english to upload for comments.

Best regards

 

[SnTMan]

acbdcl, to avoid muddying these waters any further I would like to say that in none of my previous posts did I make any mention of allowable compresive stress.

Regards,

Mike

 

[prex]

OK, let's try to put it more clearly this way (well, SnTMan, you may be right, but your way of explaining things is not really crystal clear...):

1)UG-33(f) let us calculate a single value for the required thickness of a conical section under external pressure, but that value is based on an equivalent length, that we might call more appropriately as an average length: as an example, for a small to large diameter ratio of 0.5, the equivalent length will be 3/4 the actual length of the cone. We may say as a consequence that the calculated required thickness is not really a safe value based onto the large diameter, but in some way a realistic value, that accounts for the varying diameters along the cone: it is in fact not the required thickness at the small end, but also not the required thickness at the large one.

2)Now we go to Fig.UG-28.1, where it is stated that, when the lines of support encompass different shell diameters and cones, the required thickness of each section is calculated by assuming the diameter and thickness of that section as if it was spanning the entire length between the lines of support

3)We may now apply this statement to a cone, in the simplifying assumption (often actually true) that the lines of support are at the cone transitions.

4)At the small end, we use the diameter of the connected shell (or the small diameter of the cone if there is no shell) to calculate the required thickness of a cylindrical shell spanning the total actual length of the cone (not the equivalent one!): this thickness will be smaller (how to prove this?) than the thickness calculated under 1) above

5)At the large end, we do similarly as in 4) using the large diameter: it is important to note that the resulting required thickness at the large end will be necessarily larger (this is evident) than the one calculated as in 1) for the whole cone: however, as also noted by SnTMan, this value (that could well be larger than the actual cone thickness) is not used in 1-8 calculations

6)As noted by SnTMan, the same procedure may be used at any intermediate diameter along the cone

prex

http://www.xcalcs.com

: Online engineering calculations

http://www.megamag.it

: Magnetic brakes for fun rides

http://www.levitans.com

: Air bearing pads

 

[SnTMan]

prex, well stated. Sorry for the un-clarity but this thread has ocurred over a rather long period of time, and more recently especially, I really should have been doing other things.

I also intended my remarks to be limited to the construction shown in Fig. UG-28.1(f) rather than the more general cases, my mistake for not saying so.

Hey, if this was easy, anybody could do it

Regards,

Mike