A Cost-effective Alternative to Experimentation

For full PDF click here

Project objective

Detailed analysis of a prototype inspection and maintenance tool was required to determine its effect on maintenance coolant flow when submerged in a CANDU¹ heat-transfer system. A suitable experimental apparatus that satisfied the range of tests needed would have been very costly to implement.

Computational Fluid Dynamics (CFD) was selected to perform the analysis required for this project because it offered a cost-effective, rapid alternative to experimental testing and could provide the tool designers with the required information early in the design process. ANSYS^® CFX^®was the package used to create the model.

Scope of work 

The analysis needed to determine the effect that the equipment would have on flow through a CANDU heat-transfer system. When the prototype tool is located in the system, it acts as a flow restrictor, and could reduce flow rates to unacceptable levels.

The models created had to account for the geometry of the inspection equipment, the geometry of bulk and feeder sub-channel flows using heavy water coolant, and provide results over a range of flow rates.

Work performed

CFX was ideal for use in this type of analysis because the project required that a broad range of situations be tested — applications that would normally involve a large and expensive mock-up facility. Several computer models were created, and updated, as new designs or details became available.

Flow through a header was modeled without the tool, bounded by known flow properties. The tool was then added to the header and the changes in flow properties were observed. Regions of high pressure, low flow, and recirculation were noted. A supporting leg of the tool covered an inlet to the header, and the results showed that the flow rate actually increased through the inlet. This result was unexpected, but later supported analytically. It was caused by the tool creating a region of low pressure above the inlet, forcing more fluid through the smaller nozzle; CFD software made the results very clear and presentable to observers.

The bulk flow losses were quantified as a function of header flow rate. The overall flow results also provided the location of high pressure regions, which will be examined in future work.

Results

The CFX model results clearly demonstrated that the blockage caused by the device was minimal and would not have a significant impact on operations. It was also found that the ‘worst-case’ position for the device was not where it was originally thought to be, and a more accurate ‘worst-case’ position was identified. This provided valuable information on how to better construct future models or experiments.

CFX also allowed for the generation of very clear and descriptive diagrams, which allowed for easy understanding by both analysts and management.

Direct client benefits

The client was able to see the results of the analysis much quicker and in more detail than if done by experiment. The designers were also able to see which features of their design contributed the most to flow losses, which presented an opportunity to improve the design.

The use of CFX on this project allowed for easy modifications to model geometry. The external geometry could also easily be changed to account for differences between generation plants.

¹CANDU is a registered trademark of Atomic Energy of Canada Limited.

For more information contact:

Andy Wallace, 416.207.6000 x6998, evan.davidge@kinectrics.com or

Evan Davidge 416.207.6000 x5928, evan.davidge@kinectrics.com