knuckle radius of eliptical head 1

[Jagmtr]

Hello everyone.

when we have different D/2h in ellipsoidal heads we have different crown radius by using table UG-37, sec viii div 1.

So does we have different knuckle radius for different D/2h in ellipsoidal heads by using table 1-4.4 or table 1-4.4 is only applicable with appendix 1-4f(2) & knuckle radius remain constant 0.17D for different D/2h.

Thanks in advance.

 

[SnTMan]

Jagmtr, does your head fall under the rules of Apx 1-4(f)? If so, I'd say the knuckle radius is per Table 1-4.4.

Otherwise it appears to me that knuckle radius is constant at 0.17D regardless of the ratio D/2h.

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[r6155]

@ Jagmtr
Are you calculating an ellipsoidal head or an reinforcement of opening in a ellipsoidal head (UG-37)?

Regards

 

[Jagmtr]

@r6155

I have concern only for changes in ellipsoidal profile by changing ratio D/2h.

regards,

 

[Geoff13]

Not sure I understand your question, but hopefully there will be some useful information in my reply.

Table UG-37 is a list of Code acceptable equivalent crown radii of true ellipsoidal heads of various D/2h ratios that can be used for the reinforcing calculation of suitably placed nozzles. It is not used to specify a manufacturing radius.

We need to differentiate between a torispherical head and an ellipsoidal head.
[ul]
[li]A torispherical head is formed using two separate radii. A centre crown radius (from a sphere) and a knuckle radius (from a torus). For the special geometry listed in UG-32(d) you can use that formula, but otherwise 1-4(d) is used.[/li]
[li]A true ellipsoidal does not have a crown radius or a knuckle radius. The radius changes constantly throughout the shape. For a 2:1 true ellipsoidal head you use the formula in UG-32(c), but otherwise 1-4(c) is used. UG-32(c) also allows you to manufacture a torispherical head with a knuckle radius of 0.17D and a crown radius of 0.9D, but otherwise calculate and treat it as a 2:1 true ellipsoidal head.[/li]
[/ul]

It sounds like you are trying to design / manufacture an ellipsoidal head other than 2:1 using knuckle and crown radii. If that is the case then you must use the torispherical head rules.

At that point you can choose any radii you want. If you would like a head that still closely follows an ellipsoidal head shape then the knuckle radius will not be 0.17D. I would start with the Table UG-37 crown radius and figure out the best fitting knuckle radius. However if you want to keep the 0.17D knuckle radius you can. In either case just follow 1-4(d) once you've picked your radii.

The rules in 1-4(f) are for very thin torispherical and ellipsoidal heads. They are unlikely to apply, but just check before completing your design.

 

[Jagmtr]

@Geoff13

Factor k1 is given in table UG-37 & required to calculate "tr" as mentioned in UG37(a)nomenclature. "tr" is required for reinforcement calculation so As you said, table UG-37 is used for reinforcing calculation.

but similarly when we look 1-4(a) nomenclature of "L= inside crown radius", it is given L=k1*D. here we have same k1 factor from table UG-37 to determine L for different D/2h.we have also ellipsoidal head thickness for different D/2h as per 1-4(c).

I understand that "tr" in UG37a is the same thickness as determined in 1-4(c) with some restriction for nozzle location. Please correct if am wrong.

as per 1-4(a), we are calculating L by table UG-37. If L is not manufacturing radius then what does L stand for in 1-4(a)


regards,

 

[r6155]

L is the radius of a seamless sphere, see UG-37 tr(c).

Regards

 

[JesseJ]

Per Brighton Tru-Edge...
Inside Knuckle Radius is Approximately 15% of the Diameter (See ASME Section VII, Division 1, UG-32 for clarification)

 

[Geoff13]

The L in 1-4(a) is indeed the crown radius. However it is only used in the design of a torispherical head in 1-4(d). Obviously it would also be the manufacturing radius for the torispherical crown.

The formula for a true ellipsoidal head in 1-4(c) uses only D and h for it's geometric parameters. There is no use of the L parameter, and thus L has no meaning for an ellipsoidal head. Since the perpendicular radius of an ellipse constantly changes there can't be a single crown radius during manufacturing, other than for the one special case permitted by UG-32(c).

A torispherical head is clearly much easier to manufacture and I understand why you want to do that. However I'm unclear why it seems you don't want to use the matching torispherical design formula.

 

[Jagmtr]

@Geoff13, i have understood your point.

Thank you all for your kind response.

regards,