Chapter 994 Final Preparation (New January, please give me a monthly ticket)

In Hertz's original plan, although the topic of the competition was very favorable to his side, after all, from the referees to the organizers, to the co-organizers and witnesses, all the units were from the Torch Group. What could he use to fight against them?

Especially in a competition where the scoring method is extremely complicated.

Just a small change in any detail is enough to change the overall outcome.

It's not like they haven't done this kind of thing in the past.

The experience is so rich.

Therefore, Hirsch's best expectation at the beginning was that about a dozen teams using COMSOL software would represent their schools or research institutions in the finals.

As long as you can raise your own flag in the competition held by your opponent, you will be considered a winner.

However, the final result was a big deal for him.

Two of the top three teams are our own people.

Four of the top ten teams are our own.

As for what comes next...

Sorry, the overall ranking is presented as ten groups per page.

In other words, most people will only pay attention to the first page and the page with their team.

This is no longer just a slogan.

This is like planting the flag in the enemy's base camp.

Although it can be seen that the majority of contestants choose to use TORCH Multiphysics, COMSOL has achieved a balance in the high-end games with a relatively small sample size.

To some extent, it can even be said to be an advantage.

On the contrary, it further illustrates its excellent performance in the field of CFD.

Including the group from Santa Clara University next door, although their grades were slightly lower and they were cramming to a considerable extent, they still ranked 31st overall.

It was no surprise that he represented the school in the finals.

And, it's not just that.

The competition rules do not impose too many restrictions on the calculation methods and tools used by the participating teams.

There are still about three weeks left until the finals.

This means that some teams may decide to change the computing software on the spur of the moment after seeing the results of the preliminary rounds.

Or at least be prepared.

So by the time of the finals, the number of teams in the COMSOL camp is likely to increase further...

Even... occupy the mainstream!

Thinking of this, Hirsch seemed to have seen the day when he would punch the torch and step on ANSYS, get promoted and get a raise, and reach the peak of his life.

"I think we can celebrate."

Although he tried hard to control himself, his voice still trembled a little.

Even Professor Harold, who was always serious, did not refuse this time for the first time.

On the contrary, the three students showed amazing calmness.

"I think the focus right now is to prepare for the finals."

Lyles Connor clenched his fists, obviously still not over the excitement just now, but he still tried his best to appear calm:

"A high ranking in the preliminary round will actually bring more pressure in the final."

"Moreover, this result is largely due to the problem. We can't expect to have such good luck in the finals. Therefore, we still need to increase our training on special types of problems during this period. For example, the particle flow problem in the first question of the preliminary round. If we encounter it in the finals, we won't have the opportunity to choose two out of three..."

"There is also the problem of combining computational fluid dynamics and computational structural mechanics, which is also our shortcoming in previous training. Although it is impossible to involve it in the difficulty of the preliminary round, the final is completely uncertain..."

"..."

Connor's rationality obviously affected his teammates. Hermia and George followed suit and refused to open champagne at halftime.

Of course, this can't be blamed on Hertz for being too proud.

As a representative of COMSOL, his interests are not exactly the same as those of the contestants.

For Hertz, being able to get so many teams into the finals with such high rankings is already a win.

Even if these teams do not perform well in the finals, that is an issue that the participating teams and schools need to consider.

Even if I exhaust my entire life's imagination, I can't think of how I can lose.

So in his opinion, celebrating now is not like popping champagne at halftime.

Of course, Hertz was not a person without vision.

Since someone had already refused, he didn't insist.

Instead, I started thinking about how to build momentum during this period of time.

After all, being able to gain an advantage on someone else's territory is definitely a once-in-a-lifetime opportunity that must be seized...

……

at the same time.

Compared to the jubilant Stanford, the atmosphere in China seems a little subtle.

Although the influence of this competition has not yet extended to the general public, there is no good media to report such as "This humiliating defeat has become a hot topic in the numerical computing community recently..."

But everyone in the industry is watching.

You know, in order to avoid possible intellectual property issues, many companies would buy several mainstream numerical calculation software on the market.

As long as COMSOL can truly demonstrate outstanding performance in the CFD field, there will definitely be many users willing to give it a try.

Although Torch Group has accumulated usage habits and compatibility barriers over the years and will not lose much market share immediately, for high-tech products such as computer software tools, performance is ultimately the foundation.

"Mr. Chang, should we consider making some special changes in the finals questions... based on the situation at COMSOL?"

Although Song Jingming knew that his boss had already planned this, he was not a technical person and could not understand where the turning point of this plan was, nor how exciting it could be, so he was very panicked at this time:

"I'm worried that if we can't give the other party a strong enough response, it will affect users' confidence in our products..."

Before Chang Haonan could speak, Meng Weizheng, who was in charge of the referee team, also expressed his agreement:

"Yes, Mr. Chang. Because the tie-break rules are written quite clearly, and there is a lot of information that will be made public after the game, there is a limit to what we referees can do. For example, the team that ranked first in the preliminary round had a 3-point advantage, and it is difficult to pull it back without people noticing."

“In comparison, COMSOL’s software is relatively poor at handling problems other than CFD. Can we consider setting up some multi-method linkage problems in the finals to offset their advantages in CFD?”

Hearing this, Chang Haonan frowned, fearing that the other party would be too clever to add drama to himself, so he quickly warned in a serious tone:

"I emphasize again, you judges don't need to do anything, just follow the scoring rules and ensure fair and just scoring!"

As he spoke, he knocked on the table for emphasis.

Seeing this, Meng Weizhen had no choice but to nod and promised that nothing would go wrong.

"As for the topic of the final..."

Chang Haonan paused for a moment:

"The theme of our conference is CFD, and in the finals, our every move will be magnified. If, as you said, we shift the focus of the topic to other fields in order to offset COMSOL's advantage in CFD, can't other experts participating in the conference see it?"

"But……"

Meng Weizheng obviously wanted to say something else, but was interrupted by Chang Haonan waving his hand:

“Although the specific topic of the final cannot be made public yet, I have already discussed the direction of the topic with several special guests who came to China in advance. It is about the two-phase flow calculation of polymer melts, which is 100% CFD problem, so there is no need to say more about this matter.”

In fact, Chang Haonan's original plan was to simply simplify the Falcon Z's wing optimization design problem and then move it up.

After all, it was through the engineering documents accidentally sent by Dassault that he discovered the fatal flaw in COMSOL's calculation of viscoelastic constitutive equations.

But after thinking about it carefully, he felt that the Falcon Z project involved too many commercial secrets, and even after simplification it might not be safe, so he changed his mind and "customized" a new question specifically for his opponent.

The reason for choosing two-phase flow is that the finite volume method is not good at dealing with systems with drastic changes in interface topology.

Two-phase flow is a typical example.

However, in general gas and liquid two-phase flows, highly complex viscoelastic constitutive equations are rarely used.

Therefore, he specially selected two relatively rare polymer melts as calculation objects.

A system that is not good at, with superimposed algorithmic defects.

It can be said that it was deliberately targeted.

To an outsider, this looks like a standard CFD problem.

But in fact, it can directly hit COMSOL's weak spot.

Even the weak spot that they themselves are not aware of——

Chang Haonan tried it several times with COMSOL himself, and the result was 100% divergence, and there was no possibility of obtaining a valid numerical solution.

Not only that.

If you're unlucky, even the software might crash.

Anyway, I left enough time for the camera, just waiting for the other party to perform on stage...