Starting from the ban of Matlab at Harbin Institute of Technology, a brief discussion on the current situation and development of domestic chemical engineering simulation software

Thread Content

Recently, Harbin Institute of Technology's ban on Matlab has caused a lot of discussion in the industry. It is well known that my country has many shortcomings in supporting scientific research software, and there is also a big gap with foreign countries. In this article, the author attempts to start from chemical engineering simulation software, mainly analyzing the current situation of existing chemical engineering simulation software, what impact will be had on the industry if these softwares are banned, and finally discussing how domestic simulation software can break through. What are the commonly used chemical engineering simulation software? The author here roughly divides them into four categories according to different simulation scales and applications. 1. Molecular simulation software: This type of software actually belongs to the category of chemical software. According to the simulation method, it can be roughly divided into molecular dynamics simulation and quantum chemistry simulation. Commonly used software include Gaussian, Material Studio, VASP, LAMMPS, etc. Molecular simulation software can simulate the internal bond energy of molecules, molecular spectral properties, calculate chemical reaction energy changes, etc. Molecular dynamics can also be simulated. The main application scenarios are the screening of drugs, catalysts, advanced materials, simulation of material physical properties, etc. 2. Fluid mechanics simulation software: This type of software mainly uses the finite element method or finite volume method to simulate fluid mechanics structural parts. Commonly used software includes ansys, fluent, openFOAM, COMSOL, etc. Fluid mechanics simulation can be used in the design of aircraft (aircraft, missiles, rockets), and the design of components such as stirring blades and gas distributors that are common in the chemical industry can be assisted by such software. 3. Process simulation software: This type of software simulates chemical processes through built-in models, mainly assisting chemical process design. Chemical engineering major * Distillation, extraction, heat exchange and other unit operations can be simulated with this kind of software. Current process simulation software can also dynamically simulate chemical processes and simulate real-time operating processes. Process simulation software mainly includes ASPEN, chemCAD, and Pro/II. 4.Aid design software: It is used for auxiliary design of chemical process, and can be used for drawing and piping equipment layout design. Common software for this type of software is autoCAD, PDMS, etc. At present, among these four types of software, fluid mechanics simulation, process simulation, and auxiliary design software are very mature. This maturity lies in the fact that their theories have been basically calculated and calculated. People already know which types of algorithms are suitable for which problems and what the deviations are. On the basis of the maturity of this algorithm, combined with design experience and small and medium-scale experiments, people can already use these tool calculations to design and develop products. The subsequent development and improvement space of this type of software is relatively limited, mainly optimizing algorithms, improving the human-machine interface, and introducing new modules for new processes. In terms of theory, except for some stubborn issues in individual fields (such as magnetohydrodynamics that may be involved in controlled nuclear fusion, direct simulation of turbulence, etc.), there are not many places that need to be updated in essence. Relatively speaking, the theory of molecular simulation software has not yet been perfected because it involves quantum mechanics. Taking into account that computer computing power is simplified, erroneous results often occur. There are still problems with the algorithm, and it is not very widely used at present. Although it is often used as an aid to explain experimental phenomena in scientific research, there is no case where catalysts or new materials are directly designed based on quantum mechanics. On the other hand, molecular dynamics has developed rapidly in the past 30 years and has been applied to drug site screening and pharmacokinetic research. Now many pharmaceutical companies in the West are actually using molecular dynamics for drug screening. There is indeed a big gap between us and them in this regard. Molecules are used to simulate the interaction between drugs and proteins. Traditional Chinese medicine fundamentalists often mention that the human body is very complex, and the interaction between the human body and drugs is very complex. Traditional Chinese medicine cannot be developed without relying on yin, yang and five elements. Here let’s take a look at how modern technology conducts research in the face of more complex drug and protein target effects. What impact will the ban on chemical engineering software have? Will we get stuck in chemical engineering simulation software, leading to overall backwardness in related technologies? If one day our simulation software is completely banned, how will our scientific research be affected? Frankly speaking, if the gap in chips is 10 and 100, then the gap in simulation software is the gap between 0 and 100, we * * There are almost no commercial chemical engineering simulation software. But this gap is not actually urgent. Even if our current simulation software is completely banned, it will not have a very big impact on our chemical R&D industry. First of all, most chemical engineering simulation software has cracked versions that can be used offline. If we develop product-oriented, it will not be affected at all. For example, if we use pirated fluid dynamics software to design a missile, and a war breaks out and the missile is launched, the Americans cannot say, "This missile contains American technology, and we need to impose sanctions on it," and the missile will fly backwards into the silo. Our students and researchers study these software * It won't be affected. You learned to design distillation towers and reactors by playing pirated Aspen. It's impossible for a group of Americans to rush to your house and scan your brain with a machine, and you will forget these technologies. As far as the author knows, a certain type of fighter jet used a large amount of foreign commercial software during the development process. According to the characteristics of the United States, is this software genuine or pirated? Secondly, we must correct our attitude towards computer simulation technology and be product-oriented. Some people say that although pirated versions of these software can be used, genuine software must be used to publish papers. As for the issue of publishing papers, I think the paper orientation itself needs to be changed, especially papers that focus on computer simulation applications and should be restricted. Why do we use simulation software? To reduce R&D costs, speed up R&D progress and make better products. Products are our purpose. If the purpose is to publish a paper, it is putting the cart before the horse. And as far as I know, many so-called simulations are made when the experimental results are already available. Using software to adjust parameters to piece together a simulation result that is close to the experimental results is actually meaningless. at present * * Regarding scientific research resources, the evaluation of professional titles of scientific and educational personnel, and the issuance of academic qualifications are increasingly inclined to ignore culture and use actual scientific research results as the evaluation standard. Therefore, the restrictions on related software also allow scientific researchers to adjust their concepts, doing more practical work and less writing papers. In addition, chemical engineering simulation software is highly replaceable, and relevant commercial software has alternatives. This mainly refers to process simulation software and fluid mechanics software. At present, open source fluid dynamics software includes openFOAM, etc., except that the human-machine interface is not very friendly, it can completely replace most ansys functions, and in some aspects it is even more powerful. In terms of process simulation, there has actually been a collective expiration of Aspen serial numbers in the past few months, which has attracted attention at that time. Here is a relevant article by @devin, which has a more detailed introduction to open source process simulation software. In addition, as mentioned above, these two types of computing software algorithms have basically taken shape and have been made public. In extreme cases, even if you rewrite it and make it domestically produced, it will not be difficult to meet basic needs. Finally, molecular simulation software currently has limited applications and can be temporarily followed. Molecular simulation technology us * * The gap with the world's advanced level is quite large, even greater than the chip and semiconductor fields that everyone is paying close attention to now. Although we have gaps in the chip and semiconductor fields, we have one or two companies corresponding to each link in the industry chain, including Huawei and SMIC. Students who graduate from this major have a career path, as the semiconductor industry is always developing. However, in terms of drug design, we do not have good enterprises, and the talents we train have nowhere to find jobs. From the enterprise level, this industry has not yet started. The gap in the field of molecular simulation is actually quite dangerous and can easily develop into a technological generation gap. In fact, we are now far behind the West in drug development, and the emergence of molecular simulation technology plays a big role in this. Just imagine, if there is a breakthrough in quantum chemical simulation abroad, they will have the technology to develop drugs, new materials, and new chemicals in a short time. This will also lead to breakthroughs in other fields, such as wall materials with controllable nuclear fusion, non-silicon-based chips, specific cancer drugs, etc. The technological gap between us and the West may be widened again, just as the Soviet Union had basically caught up with the West in industry * * , but in terms of semiconductors, the West opened up a new battlefield and quickly widened the gap, leading to the overall technological backwardness. Fortunately, there are currently no signs of such a breakthrough. According to the progress analysis of the current theoretical update, it is unlikely that such a breakthrough will occur in the past 20 years. We can still patiently adopt follow-up tactics. But once this happens, we must rely on the strength of the whole country to keep up at all costs. What is the gap between domestic chemical engineering software? We must first admit that there is a huge gap between our domestic chemical software and foreign countries. Although we have many alternatives as analyzed above, it is not a particularly urgent task to temporarily seek domestic chemical software to complete import substitution. From an industrial point of view, the semiconductor industry is indeed more urgent, and there is nothing wrong with giving priority to development. But in the future, when we have the energy, we should really try to make breakthroughs in chemical engineering software. On the one hand, it can be used as a bargaining chip to carry out technical negotiations and counterattack Europe and the United States. On the other hand, this field does have certain economic benefits. In addition, if we form technical barriers in chemical engineering software, we can also impose a technical blockade on the West. The biggest problem with domestic chemical engineering software is that the entire ecosystem of commercial software has been occupied by foreign software. Take process simulation software as an example: Students majoring in chemical engineering learn Aspen when they go to school, and they also use Aspen when they work. After using it for several years, if you want to switch to other software at this time, you can imagine the resistance. In addition, aspen is not a simple calculation tool. Its own database is also very powerful. It also cooperates with a large number of scientific research institutions to update data every year. In addition, aspen's large number of users also provides impetus for the company's new module development, problem updates and repairs. if we * * Although the algorithm for developing a domestic software is ready and we can make it, if we need to maintain a large database for a long time, constantly update modules, and correct bugs, it is difficult for domestic software to do so in the current software ecological environment. Another issue that cannot be ignored is piracy. Everyone loves piracy, and it has become a * Moreover, as mentioned above, we can all use pirated versions of mainstream chemical simulation software. This makes it difficult for Chinese software companies to obtain economic benefits even if they produce products similar to Aspen. In addition, copyright issues should not only be considered domestically, but also foreign piracy. The current intellectual property system favors the West because there are more foreign patents than China. But with the rise of China's scientific research industry, one day we will inevitably * * The patents held will be better than those in the United States * * At this time, with the Western style, they may overturn the table and refuse to recognize the existing patent system. In this way, the patents or software copyrights we hold in the future will become a burden to us just like the U.S. debt we hold today. As the chemical process simulation software aspen has occupied an important position in my country's chemical engineering teaching and design applications, when talking about process simulation, aspen must be mentioned. How can domestic chemical engineering software break through? Our chemical software industry can completely follow suit at this stage, but in the future, we will inevitably face the problem of overtaking and leading in this field. Overtaking means that we must have our own chemical simulation software and at the same time suppress the West. * * The market and earnings of software companies. Leading refers to how to use our leading advantages to suppress Western scientific progress and create barriers to gain profits after our chemical engineering software is in a leading position. Before talking about this issue, we have to say that if we * * What are the advantages of making a chemical engineering simulation software. The first is the language advantage. We * * Engineering staff generally use Chinese, and most foreign software has English interfaces. * There are certain language difficulties in using it. A software that is all in Chinese is relatively easy to obtain domestic support, especially in college teaching. Secondly, we have the advantage of being a latecomer in terms of programming languages. Western chemical engineering software was produced earlier and generally uses C or FORTRAN. Such languages ​​​​have already lagged behind. But from a development perspective, the updates of these software still need to be completed in such old languages, which leads to problems of computing efficiency and compatibility. Students who have installed aspen have more or less experienced installation problems, which is a manifestation of this situation. If future domestic chemical engineering software can be written in new programming languages, it will have great advantages in terms of hardware compatibility. Finally, we * * The complete range of chemical production categories facilitates the collection of data for unique processes. For example, most of the production of fine chemical products such as dyes is completed in my country. If we develop a process simulation software for this type of process, we can easily obtain data that even aspen cannot obtain. Recognizing these advantages and disadvantages, the author believes that the breakthrough direction of domestic chemical software should include the following: 1. * * Leading open source software projects First of all, with the current status of intellectual property protection in my country, the benefits from commercial software development are very limited and unsustainable. At the same time, the intellectual property protection system itself is dominated by the West and is used to suppress latecomers. * * of. so we * * It is very difficult for commercial software companies like Ansys and Aspen to emerge. Therefore, chemical engineering software development should take the open source route, which means it is free for users. As for the interests of developers, it can be determined by * * Take the lead in subsidizing it to ensure the enthusiasm of developers. At the same time, a competition mechanism can also be introduced to ensure the quality of software development. In addition, for the development results of open source software, * * You can also intervene to target specific foreign countries by blocking regional IPs or other technical means. * * Make a blockade. 2. Special simulation software for specific processes. Due to the current widespread use of pirated software, it is impossible for our domestic software to be free as a selling point, and must surpass foreign commercial software in content. Targeting specific processes is one of the most effective methods. Anyone who has used aspen and fluent will probably encounter this situation. Some processes cannot be done with these software. For example, in aspen, the reactor simulation module function is very limited, and a considerable part of the reactors cannot be simulated. For specific processes such as fuel chemical industry, coal chemical industry, etc., there is very little relevant physical property data. Moreover, the software itself is very large, but 80% of its functions will not be used by many people in their lifetime. These problems give us the opportunity to develop special simulation software for specific processes such as dye chemicals, etc., because we * * The related industries are relatively mature, and there are many ways to find cooperation opportunities to obtain specialized dye physical property data. A simulation software was specially developed for the dye production process to meet the specific needs of this industry. In the later stage, we can also cooperate and bind with the entire industry. As soon as new dye varieties and new equipment are available, they will be added to the software. In this way, in the dye segment, we can make products that can compete with Aspen. Even the development of special software can be completely led by the software user. For example, if a factory is engaged in coal chemical industry, it can make its own process simulation software based on its own products. The data source can be your own production and R&D data, and the application object can also be your own dedicated product. Simulation software can also be maintained based on its own production data and technology development progress. To put it simply, this is the route of rural areas encircling cities. Products like Aspen are large and comprehensive. We will make small but refined things to encircle and suppress Aspen in one field. 3. In-depth binding of software to the industrial chain. If our chemical engineering software development takes the route of dedicated software, we should also achieve in-depth binding to the industrial chain. From the perspective of forming technical barriers, if any patented technology is not bound to the production industry chain, then these patented technologies are equal to zero. The most obvious one is the pharmaceutical industry. In this industry, foreign countries are much more advanced than ours, and they also make a lot of money relying on patent licensing. However, because it is not tied to the industrial chain, in many cases there is nothing you can do when faced with table-turning infringement. Just like India as everyone knows, India is like this * * , the industry is so backward, but as long as foreign pharmaceutical patents are publicly not recognized, the generic drug industry can be established and huge benefits can be achieved. On the contrary, what is better is the EDA software for chip development. The development companies of these software directly cooperate with foundries and are deeply embedded in the foundry process. As a result, even if we have cracked software, we are still unable to cooperate with foundries. Our future chemical engineering software will also follow this route, cooperating with equipment design units, equipment production units, and equipment user units, and embedding it in the entire industry chain. For example, unique pumps, mixers, and even new reactors produced by Chinese manufacturers can be introduced in process simulation. In this way, even if the West plagiarizes our technology, the industrial chain will still be unusable without us. As long as we block relevant equipment manufacturers, we can also sanction foreign companies. Indian generic drugs once attracted domestic attention. In fact, it was not how advanced India’s pharmaceutical technology was, but India’s patent system that set the table against the West. 4. Chemical engineering simulation teaching should be based on open source software. Finally, I would like to make some suggestions for chemical engineering simulation teaching. Currently, all major universities use commercial software as teaching tools, which is very bad. There are several main problems. First, it creates a general environment for the use of commercial software in the industry, which is not conducive to the development of domestic software. Secondly, commercial software encapsulates the internal calculation process. Using this method for teaching can easily lead to students only knowing how to operate and not understanding the principles, which is meaningless for cultivating excellent simulation talents. In recent years, when interviewing chemical engineering graduates, many people said that they can use aspen. However, if you ask some questions about calculation principles and debugging carefully, you will find that these people only know how to operate, and only know how to fill in the numbers into the corresponding interface and then run them. They actually know nothing about the mechanism behind it. In addition, mastering open source software is also a basic requirement for practitioners in the industry. At present, the entry-level recruitment requirements for foreign CFD engineers are to master a commercial software and an open source software. For chemical engineering simulation teaching, undergraduates focus on understanding and using it. It is recommended to first learn open source software and understand the computing principles, and then use a small amount of time to get started with commercial software and experience the excellent human-computer interface of commercial software. For graduate students, it is recommended that the teaching at the doctoral level should focus on development. As mentioned earlier, we should avoid or even limit thesis writing based on simple simulation applications. If we are engaged in applications, we should focus on the final product, and calculations are only auxiliary. But we should encourage software development. If researchers develop a computing tool to address industry issues, it can be used as scientific research results to enjoy the convenience of further education, graduation, and professional title evaluation. The kernels of the openFOAM and even the Linux operating system mentioned earlier were all made by doctoral students. The fluid dynamics simulation software openFOAM was originally written by Croatian Hrvoje Jasak during his Ph.D. Has become the main open source CFD software and finally wishes us the best * * The chemical industry has its own simulation software. Of course, as practitioners, we should also understand that chemical simulation software is not very urgent. * * We should also support temporarily investing resources in more important and core areas.