Non ASME standard flange design 4

[Shashvat]

I am a trainee in the Static Equipment design of the Mechanical department. Currently, I have been provided the task of studying the applicable codes for the design of Heat Exchanger and designing a BEU type heat exchanger.

My question is that are there any guidelines or restrictions for the design of a Non standard flange ? I am currently pursuing this task, and I have designed a flange in a software, and the flange is not failing. Also, I have checked manually, and it turned out as not failing (as per ASME BPVC Section VIII division 1, appendix 2). However, I am worried about the dimensions as in I might have gone too far in selecting the dimensions for the iterative process of flange design.

 

[SnTMan]

Shashvat, other than some limitations on the hub proportions, I am not aware of any particular dimensional restrictions imposed by Apx 2. You should follow dimensions concerning bolt circle, spacing and flange OD as per TEMA Table D-5 / D5M. Other than that, your goal is to design the smallest, most compact flange that will do the job.

"However, I am worried about the dimensions as in I might have gone too far in selecting the dimensions for the iterative process of flange design."

Not sure what you mean by this statement.

Regards,

Mike

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

 

[AWDMIKE]

Ideally you should create a spreadsheet or worksheet (implying Excel or Mathcad, or other) so that you can go through the entire process and understand what impact changing certain parameters has on the overall flange. What I would also do is to have your spreadsheet / worksheet calculate the weight so that you can get a feel for how much more material (and subsequently, cost impact) you will have.

Once you are completely comfortable with how Appendix 2 works (or Div. 2, Part 4.16 which I like very much) you can proceed with using commercially available software. I always like to validate what the commercially available software calculates, or at least a thorough audit of the calculations (assuming they are clearly shown) so that I can convince myself that the design was performed in accordance with the Code.

 

[Shashvat]

Thank you SnTMan for your guidance.

Thank you AWDMIKE for your suggestion. I will follow that procedure.

 

[jtseng123]

My general advice, may not be applicable to heat exchanger:

Since I have designed many non-standard flanges from 60" to 120" for pressure vessels, the starting point is to pick a similar flange dimensions and be proportional to it. I use Taylor Forge as a go-by for flange greater than 60".

key points are:
Decide gasket type and dimensions in advance. Look for B16.5 and B16.47 series A or B to see how gasket dimensions are specified. Then come up with something similar or in the range, especially the gasket width which impacts the stress calculation.

pick the number of bolting and size that are similar to any standard. The bolting spacing is important. So take a look from any standards to see how they decide spacing relative to bolt size. You don't want large spacing that may tend to leak.

Watch out the space between bolt hole to large end of hub. It must be sufficient for nut, bolt tensioner or power tongue. You can also look from standard to see what is the space relative to bolt size. I always select the largest one to be safe.

Decide the OD of the flange. This is simple. It is relative to bolt size. So again look for standard to see the spacing and pick a reasonable one. I always select the largest one.

Flange calc is complicate with many factors to play around including hub thickness at large end/small end and hub length. Commercial program will take care of these including providing minimum spacing to hub and between bolts, and check the rigidity per Appendix 2 which is very important. Also don't forget to check stress in sitting condition. Commercial program will take care of that.

 

[r6155]

Shashvat
Begin calculation with split loose flange, without hub, the simplest and cheaper ones.

Regards
r6155

 

[BJI]

Typically heat exchanger flanges have increased thickness, exceeding the App2 calculated thickness, to allow for thermal effects. This would require a joint integrity assessment at the design stage (including the transient thermal effects by using commercial software, WRC 510 or FEA), or possibly following some rules of thumb for flange thickness increase (if validated for the application). A heat exchanger flange that has high rotation, even below the App2 rigidity check limit, will most likely have issues with thermal effects.

Not sure if you are aware but the bolt quantities should always be in multiples of 4. TEMA tables provide a reasonable starting point for bolt quantities and spacings (circ and radial). A lower number of larger diameter bolts is better than a higher number of smaller diameter bolts.

 

[Shashvat]

Thank you jtseng123, r6155, and BJI for your insights.

Jtseng123, your answer explains it all and I shall follow that path, using a spreadsheet tool and then cross verify it with a software output.

 

[tigny]

A lower number of larger diameter bolts is better than a higher number of smaller diameter bolts.

Hi BJI,

I thought the contrary was recommended, regarding flange leakage aspect. Can you explain to which aspect of the bolted flange joint assembly design your consideration applies?

best regards,

tigny

 

[Shashvat]

Hi BJI and tigny

A lower number of larger diameter bolts is better than a higher number of smaller diameter bolts.

I referred the pressure vessel design manual by Dennis Moss (3rd ed.) for flange design and found that tigny might be on the right path. To present in brief, the manual says that for low pressure equipment, the large diameter bolts would result in an increased bolt circle diameter, which would increase the moment arm, increase the total moment and hence thicker flange and hub(if considered) would be required. So for low pressure equipment, it is advisable to go for large number of smaller diameter bolts instead of a small number of larger diameter bolts. Though it would increase the assembly/disassembly time, it would take care of the leakage as the bolt spacing would be lower.

For high pressure, the area requirement for the bolts is higher, and so the bolts need to be of larger diameter. In this case, a proper analysis of the quantity and size of the bolts, bolt spacing and the bolt circle diameter needs to be done.

Please guide me if I am in the wrong understanding.

 

[tigny]

Hi Shashvat,

indeed, this is what I have read about bolted flanges assemblies: the more numerous, smaller diameter bolts, the better flange assembly.

best regards,

tigny

 

[r6155]

Same amount of corrosion has more influence in small bolt diameter.

Regards
r6155

 

[SnTMan]

Smaller bolt diameters will indeed generally yield a smaller, more compact flange, it is generally preferred. Note, not a "better" flange, but a "cheaper" one.

Most true flange design programs will start with a small bolt (typically 3/4") and determine whether a sufficient quantity of said bolts will fit the bolt circle. If not, increment the bolt size and repeat.

Regards,

Mike





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

 

[BJI]

No one is saying use over sized bolts. Relatively speaking, lower pressure flanges have smaller bolt sizes than higher pressure joints so will result in a proportionately smaller bolt circle diameters and moment arms. In my opinion, the same applies for all pressure classes. The reduction in moment arm achieved with smaller bolts is rarely warranted, since you can't adequately include enough smaller diameter bolts to justify the reduction in area. You may be able to design a slightly more compact flange per Div.1 or Div.2, but when considering in service conditions, like relaxation, thermal effects, etc, you need to increase your bolt loads significantly and run out of capacity in smaller diameter bolts. Most of the time with smaller bolts you will have to take them as close to yield as possible, after that you are out of options. For heat exchanger flanges you will usually have to increase your thickness anyway to reduce rotation caused by thermal effects, so designing the flange using the next size up bolts would typically work out better.

Yeah, assembly/disassembly time is an issue, it also can lead to less accurate assembly bolt loads and short cuts during assembly (installers performing fewer passes), but I would say the main issue is the bolt capacity for in-service loads. I have often seen heat exchanger flanges with 70-80 odd 3/4" bolts and in one case 100 3/4" bolts, which produces a really poor flange design. I guess the common misconception is that more bolts, closely spaced, with small moment arms will give more uniform gasket stress and allow a thinner flange, but it actually comes at a cost. Flanges are designed to be flexible, think about the load interaction on a typical flanged joint. A small thickness increase will better handle thermal effects, account for the increase in bolt spacing and moment arm, and permit higher bolt loads with larger bolts and therefore increase gasket stress. Excluding exotic materials, I don't think the very small cost saving on materials is justified when the exchanger will surely run into problems during operation.

There were a couple of good discussions on the LinkedIn ASME PVP forum a while ago about using more larger sized bolts, might be worth a search.