The military-industrial scientific research system of the academic master.

Chang Haonan was not idle during the time when Haojing Aviation Development Group and Luo Luo were discussing the specific introduction agreement.

In addition to dealing with the many fragmentary problems faced after the turbofan 10 was transferred to formal research and development, he also spent a lot of free time going to the 601 Institute and the 112 Factory.

On the one hand, he wanted to lick the flanker lady whom he had no chance to see in his previous life.

On the other hand, it also helps them solve some application problems of TORCH Multiphysics software.

As each trial user, especially the trial user of the Super Large Cup Professional Enhanced Edition, gradually passed through the initial learning stage, some unsatisfactory aspects of the new software were gradually discovered.

This information will be fed back to the Torch Group headquarters in Beijing. However, as the person most familiar with the TORCH Multiphysics software in the world, Chang Haonan will obviously be of great benefit to the subsequent software upgrade work if he can go to the front line to check some actual problems in person.

In the assembly workshop of Factory 112, dozens of workers are climbing up and down around the assembly frame to complete the final assembly of major components of the four J-11s.

"He Gong, are these... the second and third batches of four planes?"

According to the agreement signed with Sukhoi last year, after three batches of six large-component assembly machines are completed, Factory 112 can officially open the "complete machine assembly line" and enter the next step of localization.

"Yes, the original plan was to carry out the operations separately, but the final assembly of the first batch went very smoothly, so after communicating with experts from Sukhoi, we put the four aircraft in the second and third batches together for final assembly. "

He Ming led Chang Haonan to the plane with the fastest progress, and reached out to touch the pitot tube at the tip of the nose.

"The style of this assembly workshop seems a bit different from what I saw in the J-8C assembly workshop next door..."

Chang Haonan said and took a few steps back so that he could see the overall distribution of the aircraft's final assembly area.

Although he was not proficient in the field of production management, it could be seen that several different types of work seemed to be performed at the same time on the front, rear, left and right of the aircraft.

It's just that the coordination between them is still a bit rusty.

Including the layout of tools and equipment and the allocation of personnel, there were some unreasonable errors, which resulted in the assembly progress not looking very coherent.

Soon, a familiar feeling emerged.

"Why do I feel...this final assembly form is a bit like..."

Just when Chang Haonan was hesitating whether his judgment was correct, a familiar female voice behind him sounded:

"It's a bit like Factory 172 in Haojing?"

He turned around suddenly and found Yao Mengna wearing work clothes standing behind him.

"Sister?"

Perhaps because of his long-term stay, the other party even cut his previously long hair to ear length.

He did look a lot more capable, but Chang Haonan almost didn't recognize him for a moment.

"This final assembly model was developed by me and the team of final assembly engineers at Factory 7 when we were working on the JH-172 in Haojing. It was during the time when we were working on the JH-7, so you may feel that there are some familiar."

Yao Mengna straightened her clothes, got between Chang Haonan and He Ming, stood closer to the assembly frame, and began to introduce further:

"This method has two main features. In terms of space, it can work in parallel in multiple areas, and in terms of time, it can coordinate assembly of multiple types of work. Theoretically, the production efficiency should be at least 50% higher than that of the past continuous moving production lines. 100%.”

"Theoretically..."

As a scientific research dog, Chang Haonan quickly grasped the key points of this passage:

"In other words, the actual situation..."

Yao Mengna, who had been full of pride just now, suddenly lost her energy, put her hands on her hips, and looked angry:

"How should I put it... Assembling an aircraft from scratch is still different from previous upgrades and overhauls, and the complexity of the work involved is far higher than I expected."

"So after several rounds of adjustments, there will still be a problem where some teams are waiting for assembly tasks at full capacity, while other teams are idle waiting for assembly tasks. The overall efficiency has not changed much compared to the old method of continuous mobile production..."

In fact, the set of things Yao Mengna created is very close to the later concept of "pulsating production line".

And coincidentally, the pulsating production line in the previous life was first put into use in Haojing, and then extended to the entire aviation industry system.

However, the level of computer hardware and algorithms in the late 90s was still relatively limited, so it was very difficult to optimize this advanced production method.

"You can't say that."

He Ming, who had been silent just now, finally spoke:

"Actually, when we first used the new method to assemble the first two aircraft, the efficiency was not as high as the original continuous mobile production model."

"Now after several optimizations, it is much better, which shows that this method still has great potential. As long as it can continue to be adjusted based on actual production experience, it will definitely achieve the theoretical effect."

"After all, we have only built six aircraft. We can't rush production matters."

He Ming's comfort made Yao Mengna's mood obviously improve a lot.

"Actually, I also tried to use the differential evolution algorithm developed by you and Sister Xu Yang to optimize it from a mathematical perspective, but... the effect was not good."

Chang Haonan remembered that the last time he saw Yao Mengna, the latter still called Xu Yang Researcher Xu.

The relationship between the two seems to have improved a lot during this time...

He shook his head gently to throw away the random thoughts in his mind, and then explained:

"Although the differential evolution algorithm is advanced, it is not omnipotent. For this kind of discrete, integer, high-dimensional, large-scale, multi-constraint optimization problem, it relies heavily on the summary of problem characteristics and experience, and it is easy to fall into the trouble of local optimality. , in short, it is not suitable for application scenarios such as aircraft production lines.”

Hearing Chang Haonan's sane analysis, Yao Mengna instantly became interested:

"Teacher...Mr. Chang, do you have a better way?"

In private, Yao Mengna still used to call Junior Brother Chang Haonan, which was considered her exclusive title.

But in public, she is still relatively measured.

"Is there a better way? It's probably some path planning algorithms, but it requires some extra time to study."

"You can try using an ordinary genetic algorithm first. The effect should be better than the differential evolution algorithm."

Chang Haonan first gave some suggestions, and then added:

"My research in the field of control science and optimization algorithms is purely out of interest. Now the turbofan 10, the die forging press and the TORCH Multiphysics software all need to be handled by me personally. I am afraid that I will not have much energy to focus on these in a short period of time."

He Ming, who had been an audience next to him, was immediately stunned by this piece of Versailles.

Good guy, you just did some random research out of interest and you crushed all your peers in the country and even the world?

Do you want to let others live?

But Yao Mengna has long been accustomed to this.

She took advantage of the situation and changed the subject:

"Speaking of TORCH Multiphysics, we recently encountered some difficult problems when optimizing the structure of the J-11."

"Structural optimization?"

Chang Haonan raised his head and looked at the behemoth semi-finished J-11 in front of him:

"How far have you gone?"

Yao Mengna took him under the main wing on one side of the plane and pointed to the innermost part of the wing:

"According to the technology sharing agreement signed by both parties, Sukhoi has brought the first phase of structural optimization results to us. The difference reflected in the product is the addition of a pair of heavy loads with a maximum load capacity of 2000kg on the innermost side of the wing. pylons, increasing the total number of pylons from 10 to 12.”

"We plan to introduce this upgraded new airframe starting from the fourth batch, which is the low-knockdown assembly aircraft."

Due to structural strength limitations, the earliest Su-27 did not have a mounting point designed at this position, which was originally very suitable for mounting large-size air-to-surface ammunition.

In the previous generation, this problem was solved on the Su-27SM, but because the contract depth of Project 30 was far less than it is now, China did not use these two high-quality heavy-load mounting points until the introduction of the Su-MKK in the new century.

"Now that the problem on the wing has been solved, the problem must lie in the central lifting body itself?"

Chang Haonan probably knew the optimization process of flankers, so he asked directly.

"Yes."

Yao Mengna nodded directly:

"This is a problem that we and Sukhoi experts are struggling with. The available overload of the aircraft in a specific speed range will be reduced from +9G to +5G-+6G. Although it can be controlled by upgrading the digital fly-by-wire flight control. The impact range is minimized, but it is still a hidden danger.”

"During the structural mechanics calculation of the J-11's wing-body fusion body, we tried many methods such as dynamic overlapping grids, dynamic unstructured grids, and unstructured overlapping grids."

"But either the generation accuracy is not up to standard, or the generation efficiency is too low and it will take several months, or the calculation time step cannot be too large, otherwise it will diverge..."

"Is there such a thing?"

The grid generation algorithm of TORCH Multiphysics was specially designed by Chang Haonan.

However, the focus of the research at that time was to solve the mesh generation problem of the dirty model, and there was indeed no specific optimization for the structural mechanics/fluid mechanics composite analysis problem of this type of wing-body fusion.

When you encounter problems now, you naturally have to be very concerned:

"Come on, let's go see the specific situation."