Chapter 930 Challenge from COMSOL

Although he wanted to talk to Zheng Liangqun, he had already been transferred from Beijing to the Jiangcheng Command Post of the Air Force after commanding the 50th anniversary of the National Day parade the year before last.

At this critical juncture, it is impossible for Chang Haonan to fly thousands of kilometers to find someone just for some personal matter.

Besides, Chang Jiayao and Zheng Hui are both under 23 years old, so there is still time.

Therefore, Chang Haonan is ready to return to his job first——

But he has not forgotten the expectations that the sister units expressed for Torch Group at the aerospace-themed meeting on September 9.

In any case, the TORCH Multiphysics software must undergo a major update.

However, what Chang Haonan did not expect was that in the more than three weeks when he focused his main attention on the attack, Torch Group had already faced a strong competitor.

……

October 10th is the first working day after the National Day holiday.

Song Jingming, the office director of Torch Group, and Wei Yongming, the head of the technical department, came to the Aviation Power Group with a report.

Now that Torch Group has become a wholly-owned subsidiary of Aviation Power Group, Chang Haonan no longer has to make the trouble of making daily trips and can handle most of the specific matters at the group headquarters.

"Mr. Chang."

There was a knock on the door, causing Chang Haonan to shift his gaze from the computer screen to look towards the office door.

"Old Song... and Old Wei?"

Chang Haonan showed a look of delight, and at the same time extended his hand to signal the two to sit on the sofa:

"You two came at the right time. I was planning to contact you in two days."

He turned his office chair and said to the two people who were looking hesitant:

"Some time ago, I discussed with several core users and planned to release version 2.0 of TORCH Multiphysics at the end of this year or the beginning of next year. The most important updates of the new version are, first, the introduction of a new turbulence model based on DNS simulation and RANS simulation, and second, the introduction of the level set method for multiphase interface simulation and topology optimization problems."

"Especially the latter, which can be said to be the top priority of this upgrade. I have already pushed the theoretical part of the conservation level set method to you. I wonder how it is progressing at the application level?"

Naturally, Song Jingming had no chance to speak on this purely technical issue.

So Wei Yongming immediately took over the conversation:

"Mr. Chang, based on the algorithm theory you provided, we have implemented numerical simulations of common multiphase processes such as combustion, atomization and evaporation at a time scale of one second in our internal test version."

“As for the topology optimization problem, we are still working on it, but the progress is good. If the new version is released by the end of this year, it may be a bit tight, but it should be fine before the Spring Festival next year.”

Wei Yongming graduated from Tsinghua University and was initially responsible for developing embedded operating systems for Torch-CB Ferrari.

However, he soon showed his talents in other development fields and was transferred to the head office to be responsible for the daily work of the entire technical department, which also relieved a lot of Chang Haonan's pressure.

"Before the Spring Festival..."

Chang Haonan was naturally quite satisfied when he heard the other party's answer, but he still lowered his head and pondered for a moment:

"Old Wei, please try not to ask our colleagues to work overtime during the Spring Festival holiday. If the progress is really tight, then we can accept to launch the new version before the second quarter of next year."

Whether it is the Torch Group or the entire Aviation Power Group, due to the special nature of the work, working overtime is objectively unavoidable.

But after all, Chang Haonan had worked as a cog in the wheel for decades in his previous life, so he could always consider things from the perspective of a worker.

You know, in the field of numerical computing software, Torch Group has literally become a global leader.

I worked overtime when we were behind before, and now that we are ahead, if I still work overtime, wouldn’t it be a waste of time?

Therefore, in the computer algorithm-related industry, although Torch Group does not have the best treatment, its work intensity has always been relatively moderate.

Wei Yongming, who is directly in charge of the technical department, also agrees with Chang Haonan's ideas in this regard.

But this time, he did not nod immediately, but hesitated for a moment, looked at Song Jingming next to him, and then said solemnly:

"Mr. Chang, there is a situation now. You are not in the company during this period, so you may not know it yet."

"Oh?"

Chang Haonan raised his eyebrows:

"Tell me more specifically?"

"Just last month, two days before the National Day holiday, COMSOL of Sweden suddenly announced that they would launch the 10 version update of COMSOL MUltiphysics in late October."

Wei Yongming stood up and put the report in front of Chang Haonan:

"The details are written in this report. Please review it."

Chang Haonan immediately sat up straight and opened the folder on the desk.

In fact, he was inspired by the products of this Swedish company to name his numerical calculation software TORCH Multiphysics.

Of course it was inspired by the previous life.

However, by some strange coincidence, on this timeline, the other party became the challenger catching up from behind.

However, in the past few years, many friendly competitors, including COMSOL and ANSYS, have not been able to truly shake Torch Group's dominant position.

The content of the report is not complicated to summarize. The Swedes invited a very capable American professor to refer to the unstructured grid generation algorithm proposed by Chang Haonan a few years ago and developed a Lagrangian method that can be applied to drastic interface changes.

Since this method does not involve a high-precision format but follows the traditional finite volume method, it is very universal. Most importantly, it can still guarantee a certain degree of accuracy when the quality of mesh generation is poor.

"hiss……"

Looking at the content in front of him, Chang Haonan frowned in confusion.

If the other party can really do this without using a high-precision format, it will indeed have a considerable impact on TORCH Multiphysics, or even the 2.0 version of TORCH Multiphysics.

You know, the grid generation method specially developed by Torch Group, although efficient, is only more efficient than others when the grid generation quality is the same.

In general, it still takes a long time to complete this step.

In most cases, it is impossible for users to find the optimal solution in one cycle of mesh generation and simulation calculation.

Therefore, if you can achieve 80% or even 70% of the effect with a low-quality mesh, the appeal will be very high.

The only problem is...

Is it really possible?

"Old Wei."

Chang Haonan closed the report and looked up at Wei Yongming:

"Is there any more specific technology... uh... the technology is unlikely to be made public, but a more specific theoretical derivation process would be fine?"

"The person who developed this method is Professor Anthony Jameson of Stanford University."

Wei Yongming replied:

"He has indeed recently published several papers related to the finite volume method. If you need them, I can send them to you when I get back."

"That's not necessary..."

Chang Haonan went back to the computer:

"I can check it myself..."

He quickly typed A. Jameson into the search bar and selected Stanford University in the filter.

Soon, dozens of related papers popped up.

It was not until this time that Chang Haonan vaguely remembered the name.

"Oh my god...it's actually him?"

Anthony Jameson.

Before the 90s, he was the person who made the most theoretical contributions to the CFD field.

It is no exaggeration to say that throughout the 70s and 80s, Professor Jameson's contribution to the promotion and application of CFD exceeded the sum of all other people in the world.

Although Professor Jameson's pace of producing results slowed down a bit after the mid-90s, Chang Haonan still felt somewhat pressured to be able to invite such a legendary figure.

He sorted the search results by time and soon found the article that Wei Yongming mentioned.

A finite-volume method for the Lagrange-Euler equations…

Published in The Journal of Chemical Physics.

Compared with JCAS where Chang Haonan's paper was published, JCP is definitely a top journal in the industry.

"If this is true, then the tradition of not working overtime during the Spring Festival may have to be broken..."

As Chang Haonan thought about this, he opened the PDF document that he had just downloaded to his desktop...