Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
Thin-Out of ASME Heads 1
- Thread starter keens
- Start date
-
1
- #5
SnTMan:
I think it depends on: grade, heat treatment, forming temperature, die condition (surface roughness, wear), lubrication, speed of forming... maybe something else. I don't expect my vendor to know much about it. What about yours?
Konrad
I think it depends on: grade, heat treatment, forming temperature, die condition (surface roughness, wear), lubrication, speed of forming... maybe something else. I don't expect my vendor to know much about it. What about yours?
Konrad
panduru, no I'm not that clever this early.
Konrad, if my head vendor could not tell me in general terms where the most thinning occured in his products I'd get a different one. I'm not talking to the 1/1000, but in general.
Regards,
Mike
Konrad, if my head vendor could not tell me in general terms where the most thinning occured in his products I'd get a different one. I'm not talking to the 1/1000, but in general.
Regards,
Mike
waskillywabbit
Mechanical
Generally the knuckle area is going to be the thinnest.
As a general practice we assume 1/16" for head thicknesses 1/2" thick and under and 15% for head thicknesses above 1/2".
Brian
As a general practice we assume 1/16" for head thicknesses 1/2" thick and under and 15% for head thicknesses above 1/2".
Brian
@weldtek:
Now that sets me thinking, again !
So, the heads as delivered by the head manufacturer will have the ordered thickness at wherever is the minimum? Which would mean there's nothing the designer needs to factor-in for thinning?
It is generally believed that upto ten per cent thinning allowance is to be provided at design time, or in other words, the order will be placed for a head about 10% or so thicker than required by design. What about that, than?
You can order minimum thickness, not bothering the input material. Nice to the designer. Only problem is you never know what's the actual thickness at the skirt. Might be a headache for production department. My welding engineer would probably spit in my cofee 
Konrad
Konrad
I don't mean to convey the idea that when we purchase heads we never know what the material thickness is going to be until the heads arrive. We, like most people, have the heads quoted, and nominal thickness is part of the information given in the quote.
Looking at this differently, you could cause yourself a great deal of grief if you lock your vendor in to a starting thickness, assuming that you'll end up with the minimum thickness needed by design. It just makes sense to get the head manufacturer involved at the front end, they know the process, their equipment and its limitations.
With all due respect Konrad, it doesn't seem likely that the amount of additional thickness needed to compensate for material thinning is great enough to cause welding issues. In most cases we're talking in the order of 1/8" to 1/4" of additional thickness. Occasionally we have to taper the joint, but, we're not talking about headaches anything like you're going to have if you make assumptions on material thinning that don't hold true.
Looking at this differently, you could cause yourself a great deal of grief if you lock your vendor in to a starting thickness, assuming that you'll end up with the minimum thickness needed by design. It just makes sense to get the head manufacturer involved at the front end, they know the process, their equipment and its limitations.
With all due respect Konrad, it doesn't seem likely that the amount of additional thickness needed to compensate for material thinning is great enough to cause welding issues. In most cases we're talking in the order of 1/8" to 1/4" of additional thickness. Occasionally we have to taper the joint, but, we're not talking about headaches anything like you're going to have if you make assumptions on material thinning that don't hold true.
Weldtek, What if I put 10mm nominal thickness of the head, 10mm nominal thickness of the shell and the welding engineer assures that he has 10mm qualifications. And then he gets 12mm thick head skirt. I know, somebody should think about turning it down to 10 but what if nobody did? Welding 10mm to 12mm doesn't necessarilly take turning, because the difference can be smoothed by the weld. But if 12mm thick weld is not qualified- there's the headache!
Konrad
Konrad
OK, here is how it is commonly done in the shell and tube game, at least in my part of the world. I am speaking of of 2:1 ellip heads, but other styles of head would be similar:
The design program calculates the minimum required head thickness including any corrosion, adds forming allowance and arrives at a nominal thickness. The forming allowance is based on the nominal thickness, and is something like: 1/8" up to 1" nom, 3/16" to 2" nom, and 1/4" over that. I don't know for certain but am sure these were arrived at with input from head vendors.
The designer has two choices: Go with the nominal thickness or specify the minimum thickness. The drawbacks to the former are that the head MAWP may be lower than actual, based on the head vendor's actual minimum and/or you may be buying a thicker head than is really required. Also your nominal may be in 1/16" increments, say 11/16" and everybody hates that.
The drawback to the latter is that nobody knows the nominal head thickness, but can guess pretty closely by applying the same rules the program uses, in which case it is generally more convenient to just call out the nominal.
A vendor we used all the time also published a chart showing minimum thicknesses of heads based on plate thickness and head diameter. These minimums were generally thicker than what would be indicated using the rules built into the program. So as a third alternative, a designer could specify a head by this published minimum. Of course, this is vendor specific.
It is a curious fact that one of the most important early tasks in the mechanical design is to arrrive at a head thickness and diameter that can be economcally obtained. After that, cylinder diameters and thicknesses could be confirmed, and the rest of the design could proceed with minimal chance of major redesign.
As for the welding engineer, if the design is done as above, there are no surprises for anybody.
It may seem picky to spend some extra time at the front to nail down exactly which head to buy, but after all, this is what design is all about.
Regards,
Mike
The design program calculates the minimum required head thickness including any corrosion, adds forming allowance and arrives at a nominal thickness. The forming allowance is based on the nominal thickness, and is something like: 1/8" up to 1" nom, 3/16" to 2" nom, and 1/4" over that. I don't know for certain but am sure these were arrived at with input from head vendors.
The designer has two choices: Go with the nominal thickness or specify the minimum thickness. The drawbacks to the former are that the head MAWP may be lower than actual, based on the head vendor's actual minimum and/or you may be buying a thicker head than is really required. Also your nominal may be in 1/16" increments, say 11/16" and everybody hates that.
The drawback to the latter is that nobody knows the nominal head thickness, but can guess pretty closely by applying the same rules the program uses, in which case it is generally more convenient to just call out the nominal.
A vendor we used all the time also published a chart showing minimum thicknesses of heads based on plate thickness and head diameter. These minimums were generally thicker than what would be indicated using the rules built into the program. So as a third alternative, a designer could specify a head by this published minimum. Of course, this is vendor specific.
It is a curious fact that one of the most important early tasks in the mechanical design is to arrrive at a head thickness and diameter that can be economcally obtained. After that, cylinder diameters and thicknesses could be confirmed, and the rest of the design could proceed with minimal chance of major redesign.
As for the welding engineer, if the design is done as above, there are no surprises for anybody.
It may seem picky to spend some extra time at the front to nail down exactly which head to buy, but after all, this is what design is all about.
Regards,
Mike
- Status
Similar threads
- Question
- Locked
- Question