Casting Simulation as a Tool in Concurrent Engineering 4

Neles-Jamesbury, a large manufacturer of industrial valves, uses a lot of castings in its products. Neles-Jamesbury is continuously developing the co-operation with foundries to optimise the production and quality of the components. Neles-Jamesbury has been using CastCHECK program around six months. The intent of testing the program has been to check the functionality of the program, and it has not yet been used to change the constructions.
Both of the examples are bodyhalves of a ball valve. The material of both is stainless steel. Weight of piece A is 45 kg and of piece B is 30 kg. Piece A is cast in sand mould (in Leinovalu Foundry) and B in Croning shell (in Raahe Steel Foundry). In both cases, the CastCHECK-analysis was done after the first castings had been produced.

Fig. 11. CAD model of piece A. Data transfer between the systems is done by using STL format.

Pekka Korhonen, Neles-Jamesbury, tells more about these two case studies: "The castability was studied by traditional methods and by CastCAE-simulation in the foundries, which already showed the problem areas. However, with these components, we tried to minimise the wall thickness and the amount of machined surfaces". Smaller wall-thickness and smaller number of machined surfaces have a direct effect on the mass of the casting, and further on the price of the valve. Thus, even though the simulationdidn't show immediately a completely defect-free result, we wanted to confirm the results by test castings and not promptly give up the high set objectives.
Pekka Korhonen continues: "In both cases, the CastCHECK-analysis has been done after the casting tests. However, by comparing analysis results and test castings, one can observe that, based on CastCHECK analysis, the defects could have been forecast on the locations where they appeared in the test castings."

Pekka Korhonen discloses further: "CastCHECK-analysis gives good clues about the castability of the component compared to casting simulation programs and to the real casting tests. I believe that after some practice, even the designer with little knowledge about casting technology can achieve reasonably good results and forecasts about castability of the piece with CastCHECK. With this information it is easy to estimate the castability and open the discussion with the foundry."

Page 1 | Page 2 | Page 3 | Page 4 | Figure 1 | Figure 2 | Figure 3

User Experiences from Finnish Machine Shops

CastCHECK-program has been tested in Finnish machine shops from fall 1996 on. Examples of the companies that have participated in the testing are: ABB Industry, ABB Transmit, Neles-Jamesbury, Nokia Telecommunications, Planmeca, Sisu Axles, Sisu Tractor, Tamrock Drills and Wärtsilä Diesel.

During the testing period, CastCHECK-analysis has been used as a criterion for changing the construction, as shown by the user experiences from Tamrock: "In the case of replacing the machined component with the cast component, CastCHECK showed clearly the potential problem area in the construction. The situation was visualised to the customer by an X-ray picture from CastCHECK. After this, the customer gave an approval to change the construction to be more castable."

CASE: Two Bodyhalves of Ball Valves (Neles-Jamesbury Oy)

Neles-Jamesbury, a large manufacturer of industrial valves, uses a lot of castings in its products. Neles-Jamesbury is continuously developing the co-operation with foundries to optimise the production and quality of the components. Neles-Jamesbury has been using CastCHECK program around six months. The intent of testing the program has been to check the functionality of the program, and it has not yet been used to change the constructions.
Both of the examples are bodyhalves of a ball valve. The material of both is stainless steel. Weight of piece A is 45 kg and of piece B is 30 kg. Piece A is cast in sand mould (in Leinovalu Foundry) and B in Croning shell (in Raahe Steel Foundry). In both cases, the CastCHECK-analysis was done after the first castings had been produced.

Fig. 11. CAD model of piece A. Data transfer between the systems is done by using STL format.

Pekka Korhonen, Neles-Jamesbury, tells more about these two case studies: "The castability was studied by traditional methods and by CastCAE-simulation in the foundries, which already showed the problem areas. However, with these components, we tried to minimise the wall thickness and the amount of machined surfaces". Smaller wall-thickness and smaller number of machined surfaces have a direct effect on the mass of the casting, and further on the price of the valve. Thus, even though the simulationdidn't show immediately a completely defect-free result, we wanted to confirm the results by test castings and not promptly give up the high set objectives.
Pekka Korhonen continues: "In both cases, the CastCHECK-analysis has been done after the casting tests. However, by comparing analysis results and test castings, one can observe that, based on CastCHECK analysis, the defects could have been forecast on the locations where they appeared in the test castings."

Pekka Korhonen discloses further: "CastCHECK-analysis gives good clues about the castability of the component compared to casting simulation programs and to the real casting tests. I believe that after some practice, even the designer with little knowledge about casting technology can achieve reasonably good results and forecasts about castability of the piece with CastCHECK. With this information it is easy to estimate the castability and open the discussion with the foundry."

Fig. 12. Castings after the liquid penetrant testing: In both pictures on left the piece A and on right the piece B. The defects found in the inspection are encircled.

Fig. 13. The results of the CastCHECK runs: Left piece A, right piece B. The risk of defects and microporosity is shown by black spots (large spot-large risk).

In the Neles-Jamesbury case study, the machine shop has been testing CastCHECK program to analyse the castability of the valves and the foundries have been using solidification simulation (CastCAE) to design the feeding system for those valves. The information from both programs has been used to exchange information and to agree about the design changes and production of the valves.
Both cases show clearly the potential of the castability analysis. By analysis done in machine shops, the castability can be predicted and the construction can be redesigned to be more casting-friendly, and still maintain the suitability for its actual function, already in early stages of the design.

The designing of components using casting simulation as a tool is a quite new method in machine shops. According to the current experience and knowledge, it is obvious that castability analysis done in the machine shop can be used as significant tool in the following context:

in casting-friendly component design
design tool and as "common language" in concurrent engineering
in choosing the areas of inspection at delivery
as a tool in FMEA (Failure Mode and Effect Analysis) analysis
in specifying basic data for structural analysis
in placing critical spots to non-critically loaded areas of the component
in estimating manufacturing costs
in negotiating the price of the casting with the foundry
in negotiating construction changes with the foundry or the client/end user
as a starting point for the foundry to make rigging design e.g. for pricing

* * *

Back to CASTech Home Page