Chapter 298 Multi-physics simulation
Thinking of this, Chang Haonan suddenly felt that the name Wei Yongming seemed a bit familiar.
He was very sure that this was the first time he had contact with Qinghua University since his rebirth.
Then it should be someone who has heard it in the previous life.
“Casting process testing requires a lot of financial resources and time, and as you know, these two things are what we lack most now. At present, the pioneering countries such as the United States, Japan and Russia have been researching titanium-aluminum alloys for many years. , it is difficult for us to catch up with them in a short time using traditional methods, but the decoupling methods we are currently using have basically reached a dead end, and the project team has great differences in the subsequent route selection.”
< br>
After briefly introducing the current situation and the difficulties he faced, Wei Yongming asked Chang Haonan with expectant eyes:
"So, I came here to ask you for advice. With the current technology, In terms of means, is it possible to complete the simulation of the casting process on a computer through numerical calculations like fluid mechanics or structural mechanics? ”
Although he is only a master’s student, his summary of ideas is It is clear and organized, and he is definitely deeply involved in the entire project, rather than someone who simply does it.
His question even stumped Chang Haonan for a while.
In fact, this is not the first time the latter has faced similar problems.
Yan Zhongcheng mentioned before at the celebration party of Sanmen that Liming Factory plans to use 1Cr16Ni2MoN martensitic stainless steel to produce aeroderivative gas turbine blades through hot forging, but there is a gap between the deformation, heat transfer and organizational evolution of the material. The interaction greatly limits the application of the finite element method in hot forging processes that need to fully consider the influence of microstructure.
In the discussion about the Star of the Sea radar some time ago, Guo Lin also mentioned that a big problem faced by 14 is that the large-diameter active phased array antenna on the ship is exposed to sunlight and wind. Under the influence of environmental loads such as heavy loads and salt spray erosion, the antenna array will inevitably undergo structural deformation, deteriorating the antenna radiation performance, and requiring more power to be consumed to achieve the same transmit power.
However, increasing the input power will bring greater thermal power consumption. In addition to wasting the precious electrical energy on the warship, temperature-sensitive T/R components and array power supplies will also be generated. Temperature drift in performance causes the excitation current output by the T/R component to change, further deteriorating the radiation performance of the antenna.
Although these issues may seem unrelated and involve completely different fields, the dilemmas faced by everyone are the same——
Compared with the basics, we only need to consider Aircraft aerodynamics and structural design influenced by mechanics involve more physical fields, and they have a very close influence on each other. For the current popular calculation and analysis methods, it is impossible to calculate so many physical quantities at the same time. and optimization.
In traditional calculation methods, when nonlinear conditions are introduced, the calculation needs to iterate among multiple sets of partial differential equations many times in order to obtain convergence. However, casting problems involve material nonlinearity, geometric nonlinearity, and boundary nonlinearity at the same time. Iteration The extremely low efficiency means that the calculation time needs to be measured in months or even years, and there is a high probability that the final result will be divergent.
If these physical fields are separated and calculated separately, the coupling effect is ignored, causing the simulation results to be completely unable to fit the actual situation.
As far as existing technical means are concerned, there is no solution to this problem.
So what Chang Haonan has always wanted to do during this time is to write a new, comprehensive multi-physics simulation modeling software.
But this requires very high theoretical support, at least beyond the current era.
This is also the reason why he urgently wants to upgrade the system's theoretical ability level to LV3.
After taking some time to summarize the language, Chang Haonan gave his answer in a cautiously optimistic manner:
“Traditional decoupling methods include indirect coupling that has just been developed. All methods can only deal with weak coupling problems between multiple physics fields, and strong coupling problems cannot be directly decoupled, so there is really little meaning in continuing to try in this direction."
"For example, you just said. The casting process mentioned will involve a fluid-solid coupling. Only under the assumption of small deformation, this is a weak coupling problem, because the flow channel will affect the fluid, but in turn the effect of the fluid on the flow channel can be Ignore, such a problem can be solved by decoupling."
"But the casting process does not meet this assumption..."
Wei Yongming's eyes dimmed slightly. Obviously, he hoped to continue using numerical calculations.
Because this is the only way for China to achieve technological leap forward.
"Yes, but it is not completely hopeless." Chang Haonan nodded:
"I am developing a fully coupled solution method. If it goes well, then..."< br>
If multiphysics simulation is reduced to a mathematical problem, it is actually solving a system of partial differential equations with a high degree of nonlinearity. However, since engineering only requires numerical solutions and does not require analytical solutions, the overall difficulty should be okay.
In addition, in the original timeline, FEMlab also changed its name to COMSOL Multiphysics in 2005 and officially entered the field of multi-physics simulation. The time span is 8-10 years, and LV3 level theory should be enough. cope.
What's more, engineering software always needs to collect data for version iteration. It is impossible to expect that the first release will be perfect. Even if only the most basic force and heat coupling can be made at this stage, for almost all This is a huge boost to all industries.
“If everything goes well, there may be some phased results in the second half of the year.”
He still did not choose to open the champagne at halftime.
Wei Yongming, who was a little disappointed just now, suddenly became excited:
“So, can we continue to use numerical analysis to shorten the research cycle?”
What if someone else drew this If you grow a big pie, he probably won't give it a second thought.
But who says the score?
Before coming on this trip, Wei Yongming certainly did some homework.
Judging from past experience, the "phased results" mentioned by the general counsel are generally not phased.
It is the kind that can be directly solved by putting it in one topic.
“Of course, or rather, this is the ultimate goal of my research on algorithms and control theory.”
Chang Haonan nodded with a smile.
Life sometimes has some wonderful symmetries.
For example, he just told three students from Yenching University a few days ago about something similar.
Now I am telling a student of Qinghua University again.
“If you want to do your job well, you must first sharpen your tools. Although waiting for about half a year may cost you some time, it is worth it in exchange for higher efficiency in the end.”
As a scientific research dog doing projects in his previous life, Chang Haonan chose the second half of this year as the time node, naturally taking into account the actual situation faced by Wei Yongming and others——
Scientific research projects are due to be handed over at the end of the year. summary report.
If he supports the software until next year, the project team is likely to choose to return to the traditional route under schedule pressure to ensure that the report at least has content that is easy to write.
Experiments are of course necessary, but the cost of experiments lacking theoretical guidance is indeed too high for today's China.
(End of this chapter)