Chapter 303 Testing the plasma turbulence mathematical model
Xu Chuan didn't pay too much attention to the theft of his manuscript, after all, it was part of the plan.
That evening, Zheng Hai sent back the lost manuscript.
Although Xu Chuan was a little curious about which country had taken his manuscript from him, the security personnel were still dealing with the matter, so it was not easy for him to ask further questions.
Days passed like this, and the modeling work of plasma turbulence in controllable nuclear fusion reactors was carried out bit by bit.
In mid-September, about half a month later, the raw data of sterile neutrinos analyzed jointly by NTU and Jiaotong University was completed.
Chen Zhengping immediately called Xu Chuan and handed him the analyzed data.
"The analysis data has been sent to you, but the amount of analyzed data is still relatively large. After all, this is the special collision mathematics of the LHC. Do you need our help?" In the office, Chen Zhengping looked at Xu Chuan and asked.
The discovery of sterile neutrinos, whether it is for Xuchuan or Nanda, or for China as a whole, is actually the sooner the better. Because no matter from every aspect, the direction of sterile neutrinos is in the academic world.
It greatly enhances the academic influence of individuals, schools and countries.
Xu Chuan thought for a while and said: "That's okay. In this way, I will compile a document on how to distinguish and confirm the confidence level of sterile neutrinos, and then you will process it according to the document, and the processed data will be handed over to me to prepare.
How about Dariztu?”
Chen Zhengping nodded: "Absolutely. At present, the most important thing is to increase the confidence level of sterile neutrinos from 3 sigma to more than 5 sigma as soon as possible."
Xu Chuan smiled and said, "Then I'll trouble the teacher."
Compared to last year's New Year period, he really couldn't spare much time to process the sterile neutrino data this time.
At present, it is more important for him to seize the time to complete the modeling of plasma turbulence, and it is indeed easier to leave the data analysis of sterile neutrinos to Nanda and Jiaotong University.
Chen Zhengping smiled and said: "You don't have to be so polite. In fact, we have to thank you for giving us this opportunity. After all, for other students, this is rare experimental data."
For NTU and Jiaotong University, it is also a good thing to allow physics students who have not had the opportunity to participate in CERN experimental data analysis to participate in some high-energy physics data analysis work.
After all, not every student studying theoretical physics has the opportunity to follow a tutor or apply for an internship at CERN.
More theoretical physics students often gain some experimental experience at home or elsewhere to complete their graduation thesis.
This time, the analysis data of optically sterile neutrinos is enough for a large number of physics students from Nanjing University and Jiaotong University to complete their graduation thesis. They may even take this opportunity to obtain the qualifications to attend CERN.
As Nanda and Jiaotong University began to analyze the experimental data of sterile neutrinos in depth, Xu Chuan's mathematical modeling work on plasma turbulence has gradually come to an end.
Establishing a mathematical model for plasma turbulence is difficult even if we put aside the underlying complex mathematical calculations and related theoretical foundations.
During the modeling process, Xu Chuan encountered a lot of troubles.
Some things can be solved through various searches and learning, but some questions cannot be answered at all on the Internet.
For example, he encountered difficulties when implanting "eddy current direct simulation data" into the model.
Because there are differences between the large and small scales of eddies, and both must be solved at the same time, it will be very expensive to solve all turbulence scales in numerical simulations.
Especially when used to solve high Reynolds number flows, even with a supercomputer, the calculation may not be completed in a short time.
If it's in other fields, it doesn't matter.
But for real-time control of high-temperature plasma turbulence in controllable nuclear fusion reactors, timeliness is of vital importance.
Without real-time control, it will be too late to do anything until the plasma hits the first wall.
Looking at the crude model on the screen and the half-written modeling program, Xu Chuan let out the turbid air in his chest.
To be honest, he has been troubled by this problem for a long time. Before, he could skip it and do other work first, but now he has to make up for it.
"Perhaps, I need some foreign help." Staring at the slowly rotating model on the screen, Xu Chuan shook his head slightly.
Strictly speaking, he has not comprehensively and systematically studied mathematical modeling. He can write some simple models, but it is almost impossible to solve this problem based on his own ability.
Of course, it is normal to seek help from others when encountering problems during research, and there is no shame in it.
I'm just saying, who is better to ask for help?
After thinking about it, a figure appeared in Xu Chuan's mind. After confirming it, he opened his mailbox, sorted out the problems he encountered, the required conditions, and some relevant theoretical data and foundation, and sent it.
Then he picked up his phone again, opened Weixin, found a familiar number, and sent a message.
After a while, the phone vibrated, and a prestige reply came over.
"Well, I received the email you sent. I'll take a look at it first, but I may not be able to solve it."
Xu Chuan smiled and replied: "Thank you for your hard work. I will treat you to dinner when you come back."
For work on mathematical modeling, the first thing he thought of was his senior sister Liu Jiaxin.
He returned to China in 2018, and his senior sister went to Princeton for further studies. More than a year has passed.
With her mathematical talent and learning ability, she expected to learn a lot at Princeton, which might help him solve this problem.
Putting down his mobile phone, Xu Chuan returned his attention to modeling. Although he asked someone to help him look at it, he still had to study it himself.
After all, you can never be afraid of having too much knowledge, you are just afraid of having too little when you use it.
This chapter is not over yet, please click on the next page to continue reading! The senior sister’s reply came much earlier than Xu Chuan expected, and in the evening, a follow-up email came back.
At the same time, he also received a message from his senior sister on Weixin on his mobile phone: "Well, the plan has been sent to your email. You can see if it is suitable. If it doesn't work, you can talk to me again."
Xu Chuan was eating when he received the prestige. After replying with a hello, he cleared the food on the plate in a few mouthfuls and hurried back to the office.
When I opened my mailbox, there was an unread email lying inside.
Clicking on the email, Xu Chuan quickly browsed it.
"According to the requirements, if you need to simulate the 'direct numerical value of eddy currents', no matter you use DNS simulation, LES splitting method, or Reynolds average simulation RANS method, you will not be able to achieve the required results."
"Because the characteristic Reynolds number in plasma turbulence engineering problems is generally high, even the maximum scale within the attached boundary layer will become very small. Even if the LES model is used, the grid scale requirements are not much weaker than DNS.
.”
"This is a core fundamental issue."
"And if you want to reduce the requirements for computer hardware and computing power, maybe you can try to use an anisotropic model near the boundary layer, such as the RANS model, and use the LES model in the area far away from the wall, through double mixing.
Completing an architecture that revitalizes higher-order models”
".Suppose the Reynolds stress term t=pR^ij "$^$" represents the Favre average, and the six-component equation has the following general form:
【pR^ij/t (pu^kR^ij)/xk=pPij pΠij peij】
"The right ends are the generation term, the pressure and strain correlation term redistribution term, the dissipation term, the diffusion term and the mass term. Among them, the generation term can accurately obtain the required function."
"Introducing the transition function F allows the RANS method to act near the boundary layer, and using the LES method far away from the boundary layer, a second-order moment model under the hybrid RANS/LES framework can be constructed:"
【R^hybridij=FR^ij (1F)R^sgsij.】
"Perhaps such a high-order moment model has the potential to accurately resolve eddy flows and meets your requirements."
"Hope I can help you a little."
In front of the computer, Xu Chuan carefully read the solution delivered by Liu Jiaxin.
Perhaps the experience of studying abroad for more than a year has affected some of her personality, or she is across the computer screen, or she happens to be in her own professional field. The confidence expressed in the lines in the email is higher than before.
too much.
Of course, it has to be said that the solution proposed in this email may really solve his problem!
Not only is it detailed, but it also lists out all the key nodes in the modeling.
And to be able to make such a detailed plan in just a few hours, it is conceivable that it is not just her who has a deep understanding and high ability in modeling.
What's more, I'm afraid that I have been working hard to solve this problem since I received his email. Otherwise, it would take a few hours to come up with such a plan, with detailed nodes attached, which would be impossible to do.
After reading the email, Xu Chuan Lianshan quickly replied with a smile: "I've received it. I'll try it first. Whether it succeeds or fails, I will tell you."
"I have a hunch that it can help me solve this problem. Thank you for your hard work."
After downloading the plan from his mailbox, Xu Chuan reopened the FloEFD modeling tool. After making a backup of the original model, he began to re-modify and write the mathematical model along the lines of his senior sister's ideas.
This time, the modeling process went quite smoothly.
When he typed the last line of calculations on the computer and carefully saved the model, he relaxed.
After heaving a sigh of relief, Xu Chuan put down the ballpoint pen on his fingertips and looked at the mathematical model on the display screen with a smile on his face.
“Finally done!”
After stretching and moving his muscles, Xu Chuan checked the mathematical model he had made. After confirming that there were no problems, he copied it.
He walked into the office with the solid-state drive containing the model.
The mathematical model has been completed, so the next step is to conduct a simulation test on it.
This requires the use of a supercomputer. To run this model, even if it is just to calculate and derive phenomenological data, the computing power required is far beyond what the human brain can achieve.
Unless his brain is a sixteen-core quantum computer, this is simply impossible.
As a top 985 university, Nanjing University of Science and Technology has its own supercomputer.
NTU's high-performance computing center uses IBM's large-scale blade cluster, with a theoretical computing peak of 873.6 trillion times per second, and provides computing services for NTU's physics, chemistry, atmosphere, astronomy, materials, electronics and other departments.
Although its performance is no longer excellent now, but it 'should' be possible to test the mathematical model in his hand?
Of course, Xu Chuan is actually not sure whether the supercomputer at NTU can carry out his model, because calculations involving turbulence have always been the most complex of all calculations and require the most computing power.
Even if he has made the model as simple as possible, the test for supercomputing is probably quite large.
The theoretical calculation peak value is 34 trillion times/second. This computing power is really not much for a mathematical model of image-only turbulence simulation control. It can even be said to be pitifully small.
Maybe he should go directly to a large national supercomputing center to do this test, but in comparison, that would be a lot more troublesome, and he might even have to wait in line.
But it was different here at Nanda. As long as he needed it, it was just a matter of words to stop all other calculations and serve him exclusively, so he planned to give it a try at Nanda first.
Give it a try, you won't get pregnant anyway.
With the model in hand, Xu Chuan found the teacher in charge of managing the supercomputing center in the administrative center. After filling out an application form for the use of supercomputing, the teacher in charge of the approval stamped the application form with a red stamp.
This person's application form will be approved 100% of the time as long as he doesn't want to use a supercomputer to play games.
What's more, the teacher responsible for stamping took a look at the application content.
"Testing mathematical models of plasma turbulence."
Although he is not a professional, this administrative teacher still knows the name of plasma turbulence. This is a world-famous problem.
In this case, he passed the report without even reporting it to the leader.
I'm afraid of delaying this person's research.
After all, the report can be made up afterwards, but if it delays the boss's affairs, I guess the principal can hang him at the school gate and beat him.