Dear Richard and Members,

I am getting some discussions with the local spring hanger suppliers. I need to know how CAESAR II considers the ancillary weight in the selection of spring hangers.
1. Does CII take the ancillary weight into consideration in the Hanger Report for the nominated spring hanger tables in the setup?
2. If does, how has it been considered without knowing the type of the support? What is its relationship with operating and installed loads in the analysis?
3. If spring hangers are carrying an additional structural members what we should do? I have checked this by adding vertical support force at the same node; It increases the operating load of the hanger, but does not change the installation load at all, why?
I guess that CII does not see this force in the calculation of the installation load of the spring hangers. Is there any way to introduce this force into CII for the calculation of the installation load? Do you thing introducing mass may solve this problem? If so, how can we introduce a mass ( without adding a rigid ) into a static analysis?

Thanks in advance.

Ibrahim Demir
Design Engineer
idemir@cbh.net.au

Dear Sir,

1. CII will not automatically add the weight of the ancillary hanger components since, as you stated, the software does not know the specific arrangement/type of the spring. The user (or spring vendor) shall consider this weight.

2. If you use F1, F2, ... forces anywhere in your model you MUST include these forces in your load cases and especially in the hanger load cases; anywhere that "W" is specified.

3. Usually the spring vendor adds the hardware weight directly attached to the hanger rods to the loads supplied on the data sheets.

I remember that some vendors (e.g., Lisega) state both the analysis (data sheet) load and the SET load on their vendor drawings. Of course the vendor will not be aware of any pipe attachments such as trunnions, lugs, etc. unless specically directed by the designer.

Regards,

Bob Zimmerman
(spending the holidays on an all expenses paid trip in beautiful Al Jubail, Saudi Arabia)

It's true that Caesar does not consider any weight of spring components in it's spring design. Most company standards that I'm familiar with have some kind of note on the data sheet that the weight of spring hanger components is to be considered by the vendor.

In effect, when they go to set the spring, they should zero out the weight of the components prior to applying the preload.

If you are supplying hardware for the pipe that the vendor cannot account for (such as a dummy leg) then that weight should be accounted for in the model, if it is going to be significant. In general, I usually do some rounding up on the spring hanger loads before they go on the data sheet to account for unmodeled weights (drain valves, shoes, etc.).

One thing to keep in mind as well - when you report your loads to your civil/structual engineer, you're going to need to round that number up from your calculated value to account for the weight of the spring hanger and it's hardware. For most stuff, it's only a few hundred pounds, but if your dealing with large variables and big contants, you can be talking about a few thousand pounds that your c/s guys need to be aware of.

Bob, Edward,

Thanks for your reply. We all agree that CII does not handle the ancillary weight in the analysis.
If I was analysing a large size pipes with sufficient thickness I would not worry about ancillary weight too much. However, Pipe size is DN80 ( NPS 3" ) Sch40 with a corrosion allowance of 1.5 mm. Design temperature is 415 deg C. This pipe is connected to several sootblowers that have small allowable loads and moments. Additionally, we do not have proper space ( the Client has used space without taking the sootblower piping into account )to provide more lengths to reduce loads on the sootblowers.
We have designed the pipe and selected hanger spring rates. Everything was OK. However, the vendor came up with the different spring rate to consider the ancillary weight. I have checked the analysis with these rates by including the ancillary weight as forces, but all the cases, more loads have been shifted onto the rigid support which is not acceptable due to localised stresses introduced onto the pipe wall by the shear lugs. At the moment, changing the rigid support or the type of it is not feasible. However, I need to provide more analyses to see what options we have. This is the story.

My question was how to introduce a mass into the static analysis. I think, I can do this by playing with contents or insulation density or insulation diameter in a short length close to the spring support without introducing structural members. Playing with pipe density in the short lenght might be an alternative.

I do not want to use forces in the analysis to simulate the ancillary weights. Because, I started believing that the forces are not effective in the selection of hanger installation load. I agree that CII does consider forces in the operating case. However, the Hanger Summary with force application only will not be correct due to the unintroduced mass into the field installation load for the ancillary weight.

I hope Caesar II support team can overcome this sort of problems by introducing masses to the nodes to overcome this problem in the static analysis. Otherwise, we have to use some tricks which are not recommended for the investigations on the pipe in future.

Best regards and thanks for sharing your experience.

Ibrahim Demir

Probably the best way to accomplish this is with a small (say 1 inch long), vertical rigid element, at the spring location. For example, say the pipe runs 10-20-30-40-50, etc, and you want a spring at 40. Instead of putting a spring at 40, code a vertical rigid element, 1 inch long, from 40-41. Now put the spring at 41.

For the weight of the rigid element you should use your best guess in the initial run. Once the spring has been sized, then refine the weight of the rigid to reflect the actual hardware weight.

Using the "rigid" instead of a "force" ensures that any latter dynamic runs will properly include the mass of the hardware.

Dear Richard,

I have analysed piping by simulating the ancillary weights with changed densities of two short ( 25 mm each ) elements localised to the support nodes. It required some spreadsheet calculations for the density simulation of each additional support weight. It worked perfectly. Additionally , I did not change the stiffness/flexibility of the piping by this method. However, the danger is the future investigations in case something happens. So, we need to keep a proper record within the calculation about the assumptions we made.
This way, I can see the the contribution of ancilary weights in the field installation load of spring hangers as well.

Isn't it possible to introduce a mass to a node point for the purpose of static analysis? I know it is not possible with CAESAR II at the moment. However, you may be able to introduce this option to C II in the future. So, it may be used for static as well as for dynamic analysis.

We always look for more options in the softwares. However, getting them is sometimes not possible.

Best regards.

Ibrahim Demir

The short rigid element, connected at only one end is essentially a mass. It doesn't change the overall stiffness of theh model, because it is connected at only one end.

As to documenting the use and need for these elements, you can leave notes in the title page.

We will take a look at your suggestion, for possible implementation in the near future.

Well, first of all, it's disturbing to hear your vendor coming back with a different spring size becuase the weight of the hanger hardware is kicking it up. I'd take a close look at the catalog on that one as most of them have some range beyond the "design range" that you might be able to eat into and still keep your original design spring rate. I just looked up a figure 295 and 295A double bolt pipe clamp. For your 3" diameter, they're only 3 lbs! Even with the rod and eyenuts, I can' see more than 10 lbs or so of weight here. Something seems awfully fishy there.

As for your other issue of the load shifting and creating an excessive local stress on your shear lugs at a rigid support - why don't you use longer shear lugs (or more lugs, for that matter) to give better distribution of the load?

Dear Edward,

I thought the same earlier. But, the support arrangement is causing that. Pipe has a saddle which is supported by a structural member. This structural member is supported by two spring hangers. Total ancillary weight becomes 34 kg. For my interest, I added this load to the pipe by changing the density in a short distance, and C II came up with the new spring size that the Vendor has selected. But, I still had to play with the field installation load to reduce the load on the rigid support.

The shear lug was already under excessive stresses in accordance with code requirements and we increased the pipe thickness to SCH 80 locally to cope with it. But, load shifting from one support to another was increasing the load. Increasing the number of lugs, I thought, might be a solution. However, we have run out of space around the pipe and we had to use only two shear lugs. Additionally, it is called shear lug but still transfers a bending moment due to the thickness of the riser clamp and the manufacturing tolerances on them etc.... So, major problem became the bending moment not the shear load. Temperature is another problem for the lug design, and the allowable stress of the pipe material is low. Please do not forget the material A 106 B and the temperature is 415 deg C and the contents is superheated steam.

ASME B31.1 does not provide guidance for this kind of integral lug design other than the nominating allowable shear stress. Additionally, it asks for the calculation of single lug with full load due to the possibility of load shifting.
There are some other guides for the lug design. One of them is given by AS4041 which is similar to ASME I ( same as AS1228 ) rules for integral attachments. However, there is no length restriction for the lugs. This makes me a bit disturbed when I read WRC 198 and WRC448 which give a deeper approach for the lug design. However, these references have to be read with some sub-sections of ASME III Div 1 to get more understanding on the definitions and indices. Please note that these references put restriction on the length of the lugs in accordance with some constants for the applications in the calculation. Now, I am trying to find a copy of ASME III Div 1 to read those sub-sections. I believe this will not be easy in Australia without buying the code which is not an option. Additionally, the application of code compliance has been given for ASME III Div 1 not for B31.1. So, code compliance calculation will become a problem using ASME III Div 1 for the calculation to B31.1. I thought complying with ASME III Div 1 may satisfy the compliance to B31.1. What do you think?
WRC 448 ( Pages 17 and 18 )gives B1, B2, C1 and CAESAR II factors for the straight pipes only not for bends. Additionally does not define Mi given in the formula NB-9, NB-10 and NB-11. This reference is for rectangular attachment on Class 1, Class 2 and Class 3 pipes. We do not have definition of the pipe classes. Allowable stresses are given for 4 types of load levels ( Level A, B, C, D ). We have not got these load level definitions.

If you are able to inform us above required definitions please send me an e-mail below.

Thanks for sharing your knowledge and best regards.


Ibrahim Demir
idemir@cbh.net.au

Well,

Even with the arrangement you describe, 34kg seems like an awful lot of weight. Either that's a heavy structural member or there is a long span between the spring hangers.

As for the lug, I've not used the WRC bulletins you refer to. I typically use the Kellogg Method to check local attachment stresses as a first pass and then FE/PIPE NozzlePro if the Kellogg doesn't pass.

I must confess, I've never heard of a shear lug being overstressed due to the bending moment. Even though there certainly is a moment arm, for your situation, I can't see it being more than 1" or so.

I ran a few quick numbers here and, especially with bumping the wall up to sch 80, I'm looking at several hundred pounds for that connection, even considering your nearly 800°F temperature. I figure if you have standard wall for the rest of the pipe, the design pressure is limited to about 800psig.

Of course, this doesn't constitute and engineering recommendation. But I'd still be really surprised to find that your shear lug is overstressing in bending.

Dear Edward,

I was away for a couple days , I could not reply.

Please forgive me if I mislead you about where the stresses are. Basically, there was no stress problem with the lug. Problem was the excessive primary and secondary stresses on the pipe due to the lug actions.

Please note I have solved the problem by playing with the pipe route and adding spring hangers at different locations without using shear lugs.

Thanks to you all for your input. They were very helpful during the decision making process.

Ibrahim Demir

Wow! What a discussion. I hope it's not too late to add my 2¢.

Back at the beginning of this thread, you mentioned that you didn't see that CAESAR II was properly recognizing your forces when sizing your springs. Bob Z. mentioned that you need to include F1, F2 etc. in your load cases. He's absolutely correct, but what may have been over looked on your part is the Forces also need to be added to those first two load cases CAESAR uses to size the springs.

Since Force inputs can also be used for occasional loadings such as relief valve thrust loads, that you don't want to be considered in your spring sizing, CAESAR does not automaticly include them if you elect to have CAESAR set up your loads cases. You will need to add the F1, F2 etc. to both your hanger sizing load cases and your operating and weight load cases.

Let me give you a couple more pointers about CAESAR designed springs. If you have used the CAESAR default for the hanger table you want CAESAR to use, chances are it gave you the smallest size spring hanger that will do the job. Review the Hanger Table Help Facility for all of the options that will allow you to size the springs exactly like you want them, I like the "+400" which is "center the hot load". Sometimes this will cause CAESAR to give you the next larger size spring. This will help insure that your Hot Load or Cold Load is not at the bottom of the spring scale and you have a better chance of staying with the same spring when small changes occur that will affect the weight.

An ending note, excellent points made by Edward, Bob and Richard Aye.

Kevin Monroe