Advancement of Conformal Cooling channel design technology

Applied Technology 02

In the previous section, I helped you understand that Plan C enables reduction of cooling time to 1.7 seconds and is the most efficient. You should also have understood that coolant is actively guided around the heat accumulating part, and heat removal is performed with awareness of flow velocity and flow lines.
Is it possible to proceed from this ultimate state to even greater reduction of cooling time? Here I would like to devise various techniques for finding the limit point.
First, if there is leeway in the cooling/temperature regulation equipment, then it is easy to start increasing the flow rate. I tried carrying out this analysis, and the results showed almost no difference, as indicated in Fig. 21.

Fig. 20: Necessary cooling time back-calculated from average temperature near the gate

Fig. 20: Necessary cooling time back-calculated from average temperature near the gate

Fig. 21: Comparison when flow rate is varied between 120cc/sec, 180cc/sec, and 240cc/sec
     Solidification rate after 2.2sec

Fig. 21: Comparison when flow rate is varied between 120cc/sec, 180cc/sec, and 240cc/sec Solidification rate after 2.2sec

If the flow rate is actually raised, I think it should be possible to increase the cooling effect, but in this case, we have pursued cooling times up to the ultimate value of 1.7 seconds, and it can be determined that this should be regarded as the limit point for 3D cooling inserts.

Next, I tried examining whether improvement is possible by devising adjustments to the conformal cooling form. The design plan is shown in Fig. 22.

Fig. 22: Modified design plan for bringing water pipes even closer to the gate section and the
    solidification layer rate

Fig. 22: Modified design plan for bringing water pipes even closer to the gate section and the solidification layer rate

With a perfectly round water pipe cross section, there is an upper limit of approach to the gate part due to wall thickness, and thus a design plan is adopted which achieves greater closeness to the gate section by, as explained in section 1, making the shape flat while maintaining the cross-sectional area so there is no pressure loss.
In the case of this design plan, even if coolant conditions (120cc/sec) are kept as is, the solidification layer rate increases by 0.3% from 53.5% to 53.8%, and thus we were able to obtain analysis results enabling the average temperature of resin to be improved by 0.5°C from 210.2°C to 209.7°C.
The flow velocities and flow lines are compared in Fig. 24, and no particular tendencies are evident whereby a pressure loss might be produced.

Fig. 23: Comparison of resin average temperature with modified design plan

Fig. 23: Comparison of resin average temperature with modified design plan

Fig. 24: Comparison of flow velocities and flow lines with the modified design plan

Fig. 24: Comparison of flow velocities and flow lines with the modified design plan

If we put these alternatives together and make a judgment, the results are as indicated in the graph in Fig. 25. The region around 1.6 to 1.7 seconds appears to be the limit point when conformal cooling is applied to this preform molded article, and it is impossible to achieve any further dynamic reductions in cooling time.

Fig. 25: Prediction of cooling time with the modified design plan

Fig. 25: Prediction of cooling time with the modified design plan

However, the important point I want to stress here is that we can clearly say that "more stable molding can be achieved with the modified design plan (ConformalCooling3) than with ConformalCooling2 having 1.7-second specifications," and even with ConformalCooling design, it is crucial to carry out careful design from the above perspective.
Another point is that this case where cooling time is 1.6–1.7 seconds is close to the ultimate molding conditions, and with ordinary molds, there should still be many cases where multiple cooling issues can be improved.
The approach of conformal cooling design is more than just allowing design of 3D pipes rather than the straight water pipes of the conventional method, arranged far from the product surface, and we are convinced, that it is very important to proceed with design in a step-by-step, logical manner while incorporating examination and analysis, as described above.
If among my readers, there are some who have tried using conformal cooling, but didn't achieve much of an effect, I recommend diagnosing the problem based on the approaches I've mentioned.
If you do so, we would appreciate it if you sent a report to the current writer at OPM Laboratory, or to Rise Mold (Shanghai) Laser Technology Co., Ltd., our subsidiary in Shanghai.

We have extensive previous experience carrying out projects using conformal cooling design, but as an example where we have recently become more modest and painfully recognized the need to improve ourselves more, we encountered a situation where "for the WJ section insert of a diecast mold, no effects were achieved using the streamline system for which we have an extensive track record and experience." This happened because the amount of heat taken in by the mold differs by 3 to 4 times compared to a resin mold, and therefore it is completely impossible to keep up with the cooling capacity using the streamline system. It was a painful experience where coolant supplied from IN boiled before it was discharged from OUT, and there was no stabilization at all of the coolant flow rate, let alone a cooling effect.
This is an issue where the cooling insert of an engine WJ section has extremely thin wall thickness, and is the largest in terms of amount of heat intake, and the experience provided extremely useful information.
Fig. 26 shows the results when we had analysis done using MAGMA casting analysis software (made in Germany) for the water pipes using the streamline system provided in the diecast mold for an engine WJ part. It was calculated, as the analysis result, that with a flow rate of 0.2L and 1.0L/min, the temperature approaches the boiling point near OUT, and there is no effect. In an actual test, coolant flowed at a flow rate of 2.0L/min, but even so, the system could not get any traction against aluminum at close to 630°C.

Fig. 26: An example of analysis of a WJ insert for a diecast mold (casting analysis with MAGMA)

Fig. 26: An example of analysis of a WJ insert for a diecast mold (casting analysis with MAGMA)

Partly due to such experiences, it is important in conformal cooling design to select optimal techniques to suit TPO. Judging from the results, in the case of the above diecast, the answer is that "the problem cannot be handled except with the parallel type."
Also, pursuing this a little more, do you know what sort of composition the parallel type in Fig. 27 is designed with? I'd like you all to take a close look at the image, and think about it.

Fig. 27: An example of the parallel system for conformal cooling

Fig. 27: An example of the parallel system for conformal cooling

As a hint, the design is packed with various design intentions, such as:

  • The desire to secure a large amount of cooling
  • The desire to shorten water pipe flow length

If one wishes to maximize the flow rate, cooling capacity and similar factors by using a technique such as a diecast mold, there are cases where it is best to use this parallel method.
However, since there are limits on the forms that can be used, if you inquire with us regarding the question of whether not the method can be used in all cases, we will carry out optimal design and provide an insert, so please take that into consideration.
Based on the above, in this final section I would like to wrap up with an example of analysis of this parallel system.
This is a point where I would like to present an actual example of diecast, but there are agreements on confidentiality with our users, and thus I have analyzed what degree of change would actually happen if, in a plastic mold designed by our company, the streamline type conformal cooling channels were changed to the parallel system.

Fig. 28-①: Streamline system designed by our company and a parallel design system

Fig. 28-①: Streamline system designed by our company and a parallel design system

Fig. 28-②: Comparison of velocity vectors of parallel design system designed by our company

Fig. 28-②: Comparison of velocity vectors of parallel design system designed by our company

In the cavity shown in Fig. 28-①, the design uses the streamline system on the left, and the parallel system on the right. (The water pipe on the core side is left as is with the previous streamline system because it is difficult use the parallel system.)
If you look at the velocity vectors in Fig. 28-②, it is evident that the flow is stable, with little pressure loss.
A design is adopted which allows flow of 3 times the amount of coolant with the parallel system compared to the streamline system.
The product size is small, but the design also ensures the flow length is the shortest path.
The longer the flow length, the more disadvantageous it is for cooling effects, so this is a point to beware of.
Considering this situation by substituting into a common heat calculation equation (the formula given below), I believe it should it be possible to achieve an extremely concise solution for the superiority of cooling performance.

Q=mcΔt
Q: Amount of heat (J)
m: Mass (g)
c: Specific heat (cal/g・K)
Δt: Change in temperature (K)

*It is assumed that mass=mold, so this is a fixed value, and m in the above equation is assumed to be fixed.
*Specific heat is also fixed, so c too is taken to be a fixed value.

If the flow rate increases, then the flow rate of coolant increases, and thus the amount of heat removal per unit time also increases.
The temperature difference Δt should definitely be larger than the case where flow velocity is slow, so in the sense of removing the amount of heat Q from the mold, it should be advantageous for Δt to be larger.
Rather than make this complicated, it is more simple and clear to remember the basic formula we learned in high school physics, and solve mathematically.
When the analyses of cooling capacity are compared, the results are as shown in Fig. 29–30.

Fig. 29: Comparison of cooling performance (conventional method, streamline system, parallel system)
    *Resin surface temperature

Fig. 29: Comparison of cooling performance (conventional method, streamline system, parallel system) *Resin surface temperature

Fig. 30: Comparison of cooling performance (conventional method, streamline system, parallel system)
     *Resin surface temperature

Fig. 30: Comparison of cooling performance (conventional method, streamline system, parallel system) *Resin surface temperature

Fig. 29 is a comparison of resin surface temperatures. The overwhelming difference with horizontal water pipes in the conventional method is that, even with the streamline system for conformal cooling, there is a major effect of shifting from 70–99°C to 62–7°C, and thus if design is done from there using the appropriate parallel system, it is possible to go even further to 61–75°C.
The same thing can be said about the comparison of resin average temperature in Fig. 30, and the results are 107–128°C for the conventional method, 93–118 °C for the streamline system, and 91–113°C for the parallel system.
Next, even in analysis of the amount of warp, a comparison was made with the horizontal water pipes of the conventional method in the long X-direction of Fig. 31. The results showed about a 33% improvement with the streamline system, and about a 38% improvement with the parallel system.

Fig. 31: Comparison of warp suppression performance (conventional method, streamline system, parallel
    system) *Long direction

Fig. 31: Comparison of warp suppression performance (conventional method, streamline system, parallel system) *Long direction

Fig. 32: Comparison of warp suppression performance (conventional method, streamline system, parallel
    system) *Short direction

Fig. 32: Comparison of warp suppression performance (conventional method, streamline system, parallel system) *Short direction

In the short Y-direction in Fig. 32, a comparison was made with horizontal water pipes of the conventional method, and the results showed about a 63% improvement with the streamline system, and about a 70% improvement with the parallel system.
In this way, the maximum effect is obtained with forms that can be used with the parallel system, and thus this system should be actively used, but in actual practice, clogging and other issues of maintainability must be simultaneously considered, so adequate review is needed.
Our company, OPM Laboratory, is also concurrently engaged in development of function implementation to enable automatic design on a well-known CAD system of conformal cooling channel design techniques, and we want to put in place the infrastructure to enable anyone to do this design quickly, and achieve the maximum effect.
Also, the maintainability of conformal cooling channels needs to be improved, and we will make preparations with the intention of developing means for improving the surface roughness of the interior surface, and peripheral equipment such as techniques for cleaning to prevent the occurrence of clogging.

Kazuho Morimoto
Representative Director
OPM Laboratory Co., Ltd.