In addition to Weibo, there is also WeChat
Please pay attention
WeChat public account
Shulou
2025-01-14 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > IT Information >
Share
Shulou(Shulou.com)11/24 Report--
This article comes from the official account of Wechat: knowledge Automation (ID:zhishipai), the original title: "Lin Xueping qu Kaifeng | civilian simulation industry software breakthrough company", author: Lin Xueping qu Kaifeng
The three stages of arithmetic
The theoretical basis of computer-aided engineering CAE originated from the finite element analysis in the 1940s. From 1960 to 1970, the CAE software was still in the exploratory period, and the analysis objects were mainly the strength and stiffness of large-scale equipment such as aerospace and nuclear power. At this time, CAE can only run on Unix workstations. This is an era of poor numeracy. In order to save resources, even the display and memory are often two machines. The sheer size of computer hardware is daunting, and it is common for a computer to fill a room. Software is just an accessory to computer hardware.
The earliest generation of simulation software around mainframes include Nastran from NASA, ANSYS from Westinghouse, ADINA from MIT and Abaqus from Brown University (like ANSYS, the first customer is Westinghouse). As an expensive and scarce resource, workstation is also an embodiment of concentrated computing power. People can only go to the designated computer room and wait in line for the calculated time. Scientists who are familiar with the equations, through the combination of the command line, complete the simulation analysis in a specific scene. Since the threshold for the use of software is so high, and the neck is pinched in the hands of hardware manufacturers, its development momentum is naturally greatly reduced. Software is like a secret treasure, a small number of skilled scientists and extraordinary engineers, through unknown holes and tunnels, can get close to it.
Since IBM launched the first real PC machine in 1981, the wave of desktop PC has come. Compatible computers have become popular, and computing power has been improved. Software has also begun to be popularized, and the barriers to use have been greatly reduced. During this period, a lot of software that was closely tied to the mainframe was eliminated one by one, and independent software came to the fore. The graphical interface of the Windows95 operating system sweeps all possible obstacles. In this process, both the design of CAD software and the simulation CAE software manufacturers are busy changing their posture and embracing the era of personal computer. The design software AutoCAD has been successfully transformed from the Unix system, while Solidworks has directly become the native software of Windows. And Ansys simulation software, 14 years after its birth, began to launch computers adapted to Intel 286s in 1984. The interface is more friendly and the function is easier to use. The best example of the transformation of Vermilion Bird into an ordinary home is the ADINA of the four nonlinear simulation software. It was born at MIT in 1975, but it was not until 1988 that it was commercialized for more engineers to use. A recalcitrant horse, finally there are more people, can ride it. It is at this time that Audier, currently ranked third in the world, was born with the popularity of PC. Although the world's number one Ansys was founded in 1970, it was not until 1996 that it was listed on NASDAQ by venture capital firms. Industrial software, as if from the Olympic Fairy Mountain of Zeus, fell to the folk hall. The threshold for use has been lowered, and the capital market has begun to have a new understanding of it. Once the potential of its huge user base is recognized, the embrace of capital will naturally outweigh enthusiasm.
It is easy to notice that when a technology causes drastic changes in the times, the market pattern begins to reshuffle. Shuffling involves both death and rebirth. Some companies fail, and new faces are gratifying. Just like today's electric cars, people will be surprised that even when their bodies tremble, the new generation of kings, such as Tesla and China Wei Xiaoli, are already like children roaring past the elderly, running ahead to meet the dawn of the East in jubilation.
The same dawn shines on the track of the simulation software. Since 2010, the era of cheap and super computing represented by supercomputing centers and cloud computing has come. Numeracy has become a civilized arms race, where the United States, Japan and China compete to make efforts. The problem to be solved by the supercomputing power of these supercomputing centers is to calculate the complex events with infinitely magnified multiple parameters such as nuclear explosion and weather prediction through simulation software.
Unlike the computing power of the Windows era, cloud computing and supercomputing centers seem to be back to the centralized workstation era. Only this time a distributed desktop side has been added. In terms of form, the first generation of computing power: centralized workstations; the second generation of computing power is distributed desktop applications; and the third generation of computing power, as far as software applications are concerned, has returned to centralized computing power, but at the same time coupled with distributed desktop and even mobile applications, the most common form of the latter is the App application Mini Program.
With the popularization of calculation, the threshold has been further lowered. What about the simulation software? No accident, but also ushered in a new spring.
The rise of the scrambled simulation software in the eighties and nineties of the last century has formed a big reshuffle of the existing software. In the process, many first-class designer CAD software manufacturers were ransacked and many disappeared-these brands withered because their technology was out of date and failed to be thrown off the track by the centrifugal force of technology during the transition from workstations to PC. They died between the two storms. But many CAE manufacturers have survived tenaciously. Some of their brands disappeared in the market later, for different reasons than CAD manufacturers. They are not out of date with the first generation of computing power to the second generation of computing power, but have disappeared in the last two decades due to changes in business models: bigger fish eat them.
The new era is not compatible with them for commercial reasons rather than technical reasons, but after all, they have squeezed into a new era. In a sense, the resilience of CAE vendors is due to the fact that CAE is more professional. The threshold for its use is still too high. Generally, the simulation software is a single physical field, which only calculates structural mechanics or electromagnetism. In order to make complex simulation results, it is necessary to calculate a variety of physical fields. Simulation engineers need to be familiar with these different physical fields and different software interfaces.
For an enterprise, product designers and simulation engineers are often two groups of people. And there are far fewer simulation engineers than designers. Designers often need only one software to complete all tasks, but simulation engineers need to switch back and forth between different simulation software, such as computational structural mechanics, computational heat, computational electromagnetism and so on. Just like the operator who turns the valve in the hydraulic station, there are more than a dozen valves in front of him, each in different ways. Since each valve needs a different dialect password, scrambling is naturally the most common thing.
However, the difficulty for simulation engineers is not that they are busy opening strange valves, knowing the language of each software, but more importantly, knowing enough about their own industry knowledge. He needs to connect the industry Know-how in the engineering task to the modeling language. It takes a lot of skill.
The simulation software once again shows the high and cold knowledge software that it shuts people out. At the extreme, it is just an advanced scientific calculator that has nothing to do with the industry. This also shows the most advanced model of industrial software profitability, that is, only do general computing platforms and modeling languages to encapsulate mathematical models; as for different industry-oriented trivial scenarios, they need to be completed by simulation engineers.
This threshold is naturally very high.
As a result, there will be a new group of strangers who try to change the difficult features of the simulation software.
The gap between giant footprints can be said that if we break through in the direction of single discipline CAE (only solving any single physical field such as mechanics, heat, electricity, magnetism, light, etc.), the road has basically been blocked by monopoly oligarchs. Large-scale simulation software has already formed a platform solution by digging deep technical trenches and acquiring software of many different physical fields.
As a result, the latecomers adopted another way of thinking, avoiding a single discipline to challenge the giants, but from the perspective of engineering application, thus opening up a gap in the seemingly unbreakable iron gate.
The COMSOL simulation software from Stockholm, Sweden, has made an attempt on the simulation module of engineering application. In essence, the essence of finite element analysis and engineering simulation is to calculate mathematical equations. COMSOL originates from Toolbox of MATLAB. COMSOL focuses on engineering modules, not disciplines. The company was founded in 1986, followed by the release of FEMLAB, its first engineering application module: structural mechanics. When you see the name, you can see that it is simply a parasite of MatLab. Of course, because of the lawsuit dispute, COMSOL later rewrote the code and cut off the contact with Matlab. Two years later, the electromagnetics module was released, followed by chemistry, heat transfer and other modules. In 2005, the product name was changed to COMSOL Multiphysics--. It can be said that the name did not really form the concept of multi-physical field until 10 years after its establishment. Subsequently, a variety of engineering modules continue to increase. If you record this growing list, it seems to be a boring current account. However, these engineering application modules are seen as mobile flag planting, then the flag planting symbolizes the rising market, such as the fuel cell and electrolytic cell plate it developed in 2020, which caters to the rising hydrogen fuel industry. With a multi-physical platform as the base, the upper part of the COMSOL is like a growing plug, and new industry markets are constantly being plugged in. Chasing hot industries, it is like a sunflower, extending in different directions.
The success of COMSOL benefits from three points, one is that the basic platform is rooted in the theory of mathematical physics. Starting from the lowest theory, it naturally shows excellent ability to solve equations, and its partial differential equation PDE module is very powerful. Secondly, COMSOL provides easy-to-use software solutions to engineering problems. All product modules in the product library are standardized interfaces. Whether it is electronic, mechanical, chemical and other engineering fields, or specific physical phenomena such as heat transfer and fluid, the same software interface is used, and the operation flow is the same. This provides another option for large CAE companies that have acquired a lot of software. Large CAE companies have to make data connection and interface change for the acquired software, and users have to switch valves repeatedly for quite a long time. More importantly, COMSOL focuses on engineering modules, in which a lot of industry knowledge is encapsulated, including dozens of application modules such as electrochemistry, plasma and so on. Many modules are often functional modules that classical CAE software is difficult to have as a general software platform. These engineering modules with industry attributes let engineers breathe a sigh of relief. Their simulation tasks have industry-specific guardrails, which is an easier tool to use. It customizes special tools for the industry.
The new breakthrough of simulation software requires careful study of the gaps left by the dragon's footprints.
The third wave calls for the application of simulation software, which is constantly simplifying. The first wave of simulation software has strong specificity and low-level characteristics. It is often a general simulation software for a single discipline, such as fluid, electromagnetism, heat, structure and so on. Users need to buy different types of subject software separately. The second wave of software is more focused on the industry. Taking the engineering application module of COMSOL as a typical example, it constructs multi-physical fields at the bottom, and combines with the industry, so that application engineers can easily face their own industry for simulation development.
But this is not the end. The more simplified CAE software continues to come. They are more focused on an industry, specifically for an industry simulation. The most typical is MotorCAD, which was established in 1998, which only aims at the development of advanced motors and provides four key functional modules: electromagnetic, thermal, mechanical and motor efficiency and performance optimization.
Easy to use, it can realize multi-physical simulation design in the range of full torque and rotational speed. Therefore, it is not surprising that it dominates the motor industry, and is used by major motor manufacturers and scientific research institutions all over the world. However, if the classical simulation software is used, a large number of functions are not needed because of the excessive focus on generality. And all the modules of the MotorCAD can be sucked dry and squeezed out, it is made for the motor.
Now that industry-specific software has proved feasible, what is the next step, where is the smaller granularity and what is the lower threshold? That is the parts that are directly oriented to the equipment, even to the equipment. For example, you can only do brushless or DC motors for a particular motor, which means that smaller App will be developed directly for personalized devices. If we say that COMSOL represents the rise of functional modules based on multi-physical fields, then MotorCAD is directly oriented to the vertical industry on the basis of functional modules. On this basis, we can build more personalized scenario applications App for specific devices, and it makes sense.
The third wave of CAE software is showing a new outline.
Graphic simulation of civilian evolution COMSOL has smelled this smell, it is becoming more "App fragmentation" of its software. In its 2014 release, it included a "App developer" and released a server that supports App distribution so that different users everywhere can run these professional applications on its servers. What pleases engineers most is that the App developer includes fully graphical programming features that allow users to create App freely. By 2018, COMSOL finally set the tone for itself in the Internet age and released the compiler. Users can convert the program developed on the parent platform into a simulation App that runs independently. At run time, there is no need to install the COMSOL software. Things are getting simpler and simpler.
COMSOL's business logic is oriented to professional segments, including more than 30 engineering application function modules, rather than general software, braving the wind and waves in the Chinese market. On this basis, users have established more than 1000 App cases. Under the shadow of ANSYS, Altair and MSC, COMSOL found his own territory.
Also in China, Beijing Yundao Smart Building also avoids the general software route of classic CAE, but builds a wide range of engineering application modules and App applications on its underlying multi-physical field simulation platform, providing a lower threshold for the rising manufacturing in China.
If the software represented by the first generation of ANSYS and Nastran is highly difficult to use, it is like the 127th floor of the Shanghai Central Building. So from here, the engineering application module of COMSOL is only 20 stories high from the ground. Now the App applications of COMSOL and Cloud Road will make it at layer 10. It is getting closer and closer to the earth, and people can see it.
With the popularity of supercomputing and cloud computing capabilities, more convenient ways are emerging. If the development platform can also be developed for industry users, allowing users to develop their own App, then these applications will appear on the first floor, within reach. Cloud Road is adopting a more radical cloud approach, sinking to this position, opening up the multi-physical field simulation platform, and joint development or third-party engineering module and App development. Cloud Road has developed the electronic heat dissipation simulation software Simetherm based on the general underlying platform, which has partially replaced the mainstream commercial software Flotherm and ICEPAK; a professional consulting company in Shanghai specializing in lightweight aviation composite materials is directly calling this underlying platform to develop industry modules for domestic aircraft factories. The knowledge that used to be distributed in the hands of different people is like tangyuan spread out one by one, with all kinds of sesame stuffing and five kernel stuffing scattered and mixed together. The more civilian simulation has never been so close to the public.
Note: the market of first-generation computing power and first-generation simulation industrial software, from the global perspective, looks like a mature adult, entering a highly stable structure. This stability is determined by oligopolies, which can almost "close their eyes and order" and gobble up innovative start-ups through mergers and acquisitions. But it is different in the Chinese market, which seems to be just beginning to show a place where teenagers can stretch their muscles. How to choose the breakthrough direction of CAE software is a very strategic consideration. Adopting SaaS software as a service is one of the routes. Domestic Shanghai Siqiao, Beijing Lanwei and Germany's SimScale are all adopting this route. OnScale, the simulation software of SaaS, which is only five years old in the same camp, was acquired by ANSYS this year, which means that this cloud direction has been recognized by established simulation manufacturers. In addition to taking the route of SaaS, Yundao has further introduced the way of linkage between engineering application module and App, using the PaaS of "platform as a service", which is closely combined with industry applications. It greatly reduces the threshold of users and makes civilian simulation possible. Sweden's COMSOL has proved its feasibility. At present, many emerging industries in China, such as lithium battery and silicon carbide, have put forward new scene requirements, and there are no mature engineering modules in foreign simulation software. If you can further reduce the threshold for use, domestic CAE software can be greatly developed.
In 2015, at the NAFEMS Symposium of the International Association of Engineering Modeling, Analysis and Simulation Community, industry experts put forward the topic of simulation popularization and democratization, which became the forerunner of international advocacy of "inclusive simulation". The wide application of App based on supercomputing / cloud computing and specialized simulation will lead to the inclusive revolution of simulation technology.
With the help of the third computing wave, the third generation of simulation software may also be able to jump along with it. This time, the popularity of computing power, the simulation of universal benefits.
A brief introduction to Lin Xueping: qu Kaifeng, General Manager of Beijing Lianxun Power Consulting Co., Ltd.: founder of Beijing Yundao Intelligent Technology Co., Ltd.
Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.
Views: 0
*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.
Continue with the installation of the previous hadoop.First, install zookooper1. Decompress zookoope
"Every 5-10 years, there's a rare product, a really special, very unusual product that's the most un
© 2024 shulou.com SLNews company. All rights reserved.