Transient Thermal Simulation of Multi-Pass Valve Welding
Transient thermal FEA of a multi-pass field weld on a trunnion-mounted ball valve, checking that the soft seals stay below their temperature limit on an accelerated welding schedule.
Project brief
A valve manufacturer needed to weld extension pup pieces onto the butt-weld ends of a large, high-pressure trunnion-mounted ball valve, on a compressed welding schedule. Heat from the weld travels through the thick body wall towards the elastomer O-ring and the polymer seat inside. If either overheats, the valve has to be stripped and resealed.
Shirsh Technosolutions modelled the welding sequence as a transient thermal analysis in ANSYS Mechanical to predict how hot the seals would get, and when.
What was at risk
The procedure called for a root pass followed by several fill and cap passes, laid down quickly one after another. With little time between passes the heavy wall cannot shed its heat, so the temperature builds from pass to pass.
The seals sit some distance from the weld, which makes the risk easy to underestimate. The heat reaches them late, so the critical moment may come after the welder has finished rather than during welding. Both seal materials have a strict temperature limit, and the client needed to know whether the schedule respected it or whether cooling measures were needed.
Solution offered
We built a 3D thermal model of the valve end, the pup piece and the seal grooves, and split the weld joint into rings matching the planned pass sequence. Instead of a moving heat source, each pass was applied as a lumped thermal load over its whole ring. For components this far from the weld, that gives the same answer at a fraction of the run time.
The pass sequence was handled with element birth and death through a short APDL command block. Elements for the later passes were switched off at the start and brought back in turn, each with its weld temperature applied, at the scheduled interval.
Temperature-dependent conductivity, density and specific heat for the cast valve body and the stainless pup piece came from ANSYS Granta material data. The run was extended well past the end of welding to capture the cool-down, and temperature probes on the O-ring and seat grooves recorded the full time history.
Common questions
Why not model a moving torch? A moving heat source matters for distortion and residual stress close to the weld. For temperatures at components well away from it, a lumped pass gives the same heat input and solves far faster.
Why run the analysis past the end of welding? Heat keeps travelling through the wall after the arc stops. The peak at a remote seal often arrives during cool-down, so stopping at the last pass can miss it.
Challenges we overcame
Heat building up from pass to pass on a compressed welding schedule
Seal temperature peaking after welding ends, not during it
Many weld passes in one model without a moving heat source
Temperature-dependent properties for two different steels
How we executed
- 1
Split the weld joint into rings matching the planned pass sequence
- 2
Assigned temperature-dependent thermal properties to the cast body and the pup piece
- 3
Set up a multi-step transient with one step per pass and an extended cool-down
- 4
Used an APDL command block for element birth and death to bring each pass in on schedule
- 5
Placed temperature probes on the O-ring and seat grooves and tracked their full time history
Project outcome
Mapped how the heat travels through the wall and how long it takes to reach the seals.
Showed that the seal temperature peaks during cool-down, which a check of the welding window alone would miss.
Gave the client a clear check of seal temperatures against their limit for the planned schedule.
Where the margin was small, cooling measures such as heat sinks or minimum interpass hold times were checked in the model.
Project gallery
Where this lives
Software used
At a glance
We apply the same rigour to every engagement
Let us show you what's possible for your product.