| Applies To | |||
| Product(s): | AutoPIPE, | ||
| Version(s): | All | ||
| Area: | modeling | ||
| Original Author: | Bentley Technical Support Group |
Dec 2014, AutoPIPE V8i 09.06.01.10
Problem:
How to model a U-bolt (ubolt) in AutoPIPE? If the U-bolts are clamped tight can anyone tell me the proper way to model these supports? What friction factor should I use?
Solution:
A U-bolt can be modeled as using one of the following techniques:
1. Insert a Guide support type with the appropriate settings.
OR
2. Combination of 2 or more incline type supports at the same node point
OR
3. For the most accurate analysis, model the structure as beams in AutoPIPE and attach the the pipe to the structure using a single guide support (see AutoPIPE help Modeling Approaches> Frames> Pipe Rack (Method 1)).
Notes:
1. An accepted friction factor value ranging from 0.3 to 0.4 is typically utilized for metal-to-metal contact at support locations, unless low-friction pads are incorporated, in which case a value of 0.1 may be applicable (it is advisable to consult the manufacturer for the precise friction coefficient).
2. In instances where the U-bolt is secured tightly, non-metallic padding is generally employed; otherwise, the pipe may experience denting due to hoop radial expansion, or the support may fail due to axial growth. AutoPIPE does not automatically model this denting. One would need to update the model accordingly by adding another support, suggest axial incline.
3. The majority of friction arises from the horizontal surface where the weight is applied, and thus the aforementioned factors should be considered. If the tightness is deemed excessive, it can be represented as an equivalent normal radial force (contact pressure*contact area) and that can be used to evaluate the friction factor by dividing the force by the gravity reaction.
4. When modeling multiple supports at a single node point (model approach #2 above), total combined support reaction can be wrong. It is important to note that in reality the pipe is secured with u-bolts to the structure. At these locations, if the analysis indicates movement in the lateral direction a lateral friction force maybe calculated, despite the fact that the pipe is not capable of actual lateral movement. Consequently, the summation of the restraint reactions would be inaccurate.
Example:
-
- Vertical: Incline (+Y) with Rigid Stiffness
- Lateral: Guide (up/down gaps = 2500 mm, left/right gaps = 0 mm) with 638 N/mm stiffness
During this example analysis, both axial and lateral movement were calculated. Since the Ubolt was modeled with two completely independent supports, the results will show the following loading :
1. A lateral friction force component on the vertical Incline support
2. A lateral force component on the Guide support.
Again, the summation of the restraint reactions would be inaccurate because in reality the pipe would not be allowed to move laterally relative to the vertical support structure that the pipe is Ubolted too. Recommend modeling using option#3 above, connect supports to modeled frame structure,