INTELLIGENT WORK FORUMS FOR ENGINEERING PROFESSIONALS
Come Join Us!
Are you an Engineering professional? Join EngTips now!
 Talk With Other Members
 Be Notified Of Responses
To Your Posts
 Keyword Search
 OneClick Access To Your
Favorite Forums
 Automated Signatures
On Your Posts
 Best Of All, It's Free!
*EngTips's functionality depends on members receiving email. By joining you are opting in to receive email.
Posting Guidelines
Promoting, selling, recruiting, coursework and thesis posting is forbidden.

WRC 107  nozzle loads(5)

asv80 (Mechanical) (OP) 
15 Mar 07 16:42 
This is totally new to me and I would request your advice.
We have been provided the allowable forces and moments for weight, seismic and thermal loads for different nozzles and asked to design the nozzles to withstand them.
When entering the values for forces and moments in WRC 107 calculation, should the nozzle be able to withstand eac h of the loads individually, the maximum of the three, or the sum of all three loads.
We get reasonable answers for the individual and the maximum of the three case. However, the stresses seem a lot when we add all three values and apply them.
Thanks in advance. 

asv80 (Mechanical) (OP) 
15 Mar 07 20:14 
I looked through some of the past threads and I think all the loads need to be applied simultaneously. However, we find some unreasonable stress values.
The options we tried were having heavier wall pipe for the nozzle and the attached shell and also reinforcing pads. These do not seem to have real effect and become unusually thick.
What are the design changes that are typically considered for nozzle loading applications? Does the thickness of flange have any effect?
We are going to ask our customer if the loads may be reduced. But before that we want to give them an idea of what effect the present values have in terms of cost and size.
Thanks. 

(2) jte (Mechanical) 
16 Mar 07 10:35 
asv
You don't have to combine the loads as long as they will not occur simultaneously. Since a seismic load will most likely occur when the vessel and piping are hot and since weight exists most of the time, yes, you have to consider these loadings as acting simultaneously.
WRC107 studies the stresses in the shell. Thus, increasing your nozzle wall thickness would have no/little impact on the results. This is sometimes an issue which will get you, though: You need to independently verify that the nozzle neck can handle the loads. I believe Compress does this, but you need to verify. Another issue which might get you is that Compress presumes that the loading is applied at the intersection of the nozzle and shell. If the piping loads are provided at the face of flange you'll need to translate them over to the shell. You also need to determine what your allowable stress is. Typically when combined with internal pressure you'll use a 1.5*S limit; a seismic load may come with a 1.2 multiplier on the allowable stress so consider that also.
For dealing with shell stresses: * First run the piping loads without any changes to your pressure design. You probably already have a repad and that will help. * If the stresses are high at ID of the repad, increase repad thickness and/or fillet weld size. * If the stresses are high at the OD of the repad, increase repad diameter.
Increasing shell thickness / using an insert plate is always nice from my perspective (the owner/user perspective), but often this is not economical or the shell has already been rolled. Depending on your relationship with the piping design part of the project it may be very reasonable to work with them. Often the piping loads are developed assuming the shell is rigid. If you develop spring rates and have the piping designer rerun their analysis using them you'll find the loading reduced. Sometimes sharpening the pencil is all it takes!
jt 

asv80, Quote:We have been provided the allowable forces and moments for weight, seismic and thermal loads for different nozzles and asked to design the nozzles to withstand them.
Engineering companies have different criteria for load combinations. Ask your customer for the criteria, or ask them to provide you with the worst case loads in terms of forces and moments that are consistent with the WRC 107 convention (i.e. Pr, Vc, Mc, etc.) 

asv80,
jte and doct9960 have both provided valuable information.
There are a lot of threads in several forums here at engtips, I suggest browsing through these to get a better understanding of WRC107 and what it is capable of and its limitations. Also, if you don't have a copy of the bulletin I suggest you obtain one and review it.
WRC107 provides a method to determine stresses in cylindrical and spherical shells (and by extension, formed heads) that result from loads acting on nozzles or other attachments. Two important points: this particular bulletin (1) does not address stresses due to pressure (2) nor does it address stresses in the nozzle neck, the nozzle neck is essentially treated as a rigid attachment.
Other methods of analysis of stresses in the shell and nozzle are available. Welding Research Council bulletins 297 and 497 address these issues and extend the work done for WRC107. Analysis can also be done using finite element analysis (FEA).
COMPRESS provides additional calculations to determine the stress in the nozzle neck under the combined action of internal pressure and the external loads. The nozzle neck is treated as a beamcolumn to determine the primary longitudinal stress due to axial load plus bending; the analysis also includes the effect of internal pressure.
COMPRESS includes a calculation to account for the increased local stresss in the shell at the nozzle opening due to internal pressure. This is based on published ASME paper PVP399 as referenced in the COMPRESS reports.
Because of the limitation (2) noted above increasing the nozzle neck thickness will have no effect on the stresses calculated per WRC107 (it may change the stress due to pressure as determined by PVP399). This leaves you with the option to increase the shell thickness or add a reinforcing plate. When a reinforcing plate is added COMPRESS performs the WRC107 analysis at two locations: (1) in the shell at the nozzle neck's OD, and (2) in the shell at the OD of the reinforcing plate. Be careful not to use too wide of a reinforcing plate; a plate that is too wide will be flexible and won't meet the assumption of WRC107 that the attachment is "rigid" leading to inaccurate results.
Tom Barsh Codeware Technical Support 

LSThill (Mechanical) 
17 Mar 07 5:22 
asv80 (Mechanical) You may want to look at Codeware Compress FENozzle as a reference check for WRC 107 & WRC 297. Wsing FENozzle/Nozzle Pro by Paluin Research Group www.paulin.com. 

asv80 (Mechanical) (OP) 
20 Mar 07 16:05 
Thank you everyone for the suggestions. I also went over some of the previus threads in the forum.
We do not have capabilities to do FEA for this. From the compress values for various loads, the bending moment seems to have more influence on the stress values. We are checking if that value can be reduced.
One question on combining the different loads. Is it the sum of the different loads or is there a different formula as Compress seems to require one effective value only. Like doct9960 had said are there any typicall criteria?
Thanks. 

I don't understand the question about "Compress seems to require one effective value only." Can you provide more information? The stress in the shell calculated by WRC107 method results from the effects of all applied loads and moments (P, VL, & Vc, and ML, Mc, & Mt). Each load has an effect, some more than others. The radial force and bending moments have the greatest influence on the stresses, the shear forces and torsional moment have much smaller effect (see thread1259180585). Note that each combination of loads will generally produce a different stress or distribution of stress in the shell. Thus if you are given a set of loads to be applied to the shell, whoever has specified the loads should also specify if the loads are to be applied simultaneously, if they are are reversible, etc. Tom Barsh Codeware Technical Support 

asv80 (Mechanical) (OP) 
21 Mar 07 8:37 
Tom, We were provided a table of nozzle sizes and the values of VL, Vc, ML, and Mt for three conditions  weight, seismic, and thermal. Other values are not provided and we take those forces/moments are not present.
When we supply values to compress we input all the values for one type of load and calculate the stresses where we are okay.
When all three loads are applied together, do we add the three loads (say VL = VLweight + VLseismic + VLthermal) or any other way. We did not get an effective response from the customer on how these 3 values are to be combined and they asked us to give them the maximum load we calculate.
Thanks. 

ukinam (Mechanical) 
21 Mar 07 9:01 
asv80,
Are you sure you got "allowable" loads and not "actual" loads from your customer? Many companies provide a tabulation (sometimes formula) of allowable loads for different nozzle sizes. This table of allowable loads are not usually categorized in separate weight, thermal, seismic or wind loads; they are usually a combination of different loads and they all depend on the company's criteria or practice for combination loadings. That is why it is best to ask your customer for the criteria for load combinations.
For the weight, thermal & seismic loads provided earlier to you, I assume they are in terms of 3 forces and 3 moments (i.e. Fx, Fy, Fz, Mx, My, Mz). Also, your customer should have given you the direction and orientation of those loads in terms of the x, y, z axes. In case your customer don't have any criteria, here's what you can do... 1. Consider different cases: Case 1: weight Case 2: thermal Case 3: weight + thermal Case 4: weight  thermal Case 5: weight + thermal + seismic Case 6: weight + thermal  seismic Case 7: weight  thermal + seismic Case 8: weight  thermal  seismic In the above different cases, add forces and moments in terms of the same axis. Note that the weights above are the operating weight. If the customer indicated separate operating and hydrotest weights, consider hydro weight as a separate case. 2. Input each case in Compress WRC 107 (follow the correct sign convention) and determine the worst load case. 3. From that worst case, do what jte suggested in "dealing with shell stresses". 4. If the shell stresses still come out unreasonably high, go back to your customer and ask them if they can reduce the loads. Otherwise, you have no option but to increase shell thickness.


It is common that different "sets" of loads may exist for different vessel conditions (operating, shut down, vacuum, etc). Generally, each condition must be analyzed separately. At this time, COMPRESS provides option for only a single "set" of nozzle loads. We may make provision for entry of multiple load sets in a future release of COMPRESS. You can consider different sets of nozzle loads now by saving a copy of your COMPRESS vessel file. Then change the nozzle loads and rerun the Code calculations. Tom Barsh Codeware Technical Support 

LSThill (Mechanical) 
21 Mar 07 17:19 
TomBarsh (Structural)
REGARDS TO THE ABOVE SUBJECT WRC 107  nozzle loads, between Feb.  June 1999 using codeware compress FENozzle you did send me the input excel spreed sheet format to WRC 107  nozzle loads. Please check if your still have excel spredd sheet may help other compress user's.
Stephen 

asv80 (Mechanical) (OP) 
23 Mar 07 12:31 
We were told by our customer that the seismic load need not be combined with the weight and thermal. With this condition, we do not get very high stresses.
Tom: If you still have the excel file, it would be very useful.
It was a new learing for us and thanks for all your help. 

asv80,
Your customer's instruction to not combine seismic doesn't seem right. I suggest you also run a calculation where seismic is combined with weight plus 1/2 of thermal, and see if this can be accommodated in your design. Also check the Code for requirements on loadings. 

or better yet, ask your customer for the actual piping or external loads 

asv80 (Mechanical) (OP) 
23 Mar 07 15:39 
The reason given for not including seismic load was horizontal installation. Is that correct?
Also, what we did was tell them that individual loads are okay and gave them the maximum total forces and moments the design will withstand.
Thanks.


ukinam (Mechanical) 
23 Mar 07 17:49 
asv80,
"The reason given for not including seismic load was horizontal installation. Is that correct?"
I don't quite get what you are saying in your above statement. For sure there are horizontal seismic forces and horizontal thermal forces. And how about wind loads? Even if seismic is the governing factor over wind in your vessel design, it should be considered in your customers piping stress analysis.
This is the dilemma with your customer giving you allowable loads categorized in weight, seismic, thermal, etc. There are a lot of variables or cases to be considered. 

asv80 (Mechanical) (OP) 
25 Mar 07 16:42 
ukinam,
the nozzles will be on a vessel that is horizontally mounted. Per the customer, seismic loads are applicable only for vertically mounted units. I asked another person here and he too says that seismic analysis will not be required for horizontal cases. Also, Compress too seems do the same. It gives a message like seismic loads will be applied only in case of vertical vessels. Can some experienced users shed some light if this is the case and correct?
Thanks. 

asv80, COMPRESS applies seismic and wind loads to horizontal vessels (saddles) if these codes are specified. Of course, how you decide to apply the piping loads to nozzles is another issue. Tom Barsh Codeware Technical Support 

cycguy (Mechanical) 
26 Mar 07 14:46 
Gentlemen,
It is also important to point out that in Compress, that the WRC loads applied to the nozzles do not get transfered to the analysis of the vessel. The WRC analysis only checks the nozzle to shell junction for stresses. The loads should also be applied to the vessel via the "Loads" button in the toolbar. You can input vertical loads with an eccentricity and lateral loads. This is for checking the vessel support. Correct me if I'm wrong TomBarsch.
Thanks. 



