Вы находитесь на старой версии сайта. Перейти на новую: technospark.ru
Oleg Lysak, TechnoSpark: “Market will change when 3D-printing becomes as clear a service as CNC-machining”

“There is no market in additive technologies in Russia yet. So market rules are not applicable. Rather, in most cases they are R&D efforts that are taken in the let’s-try-it-here-let’s-try-it-there fashion. There is only one tiny market segment that really exists, and it is dentistry. Additive technologies haven’t got a large share in the world market either. Their global market is worth $8 bln despite all the hype (aviation, space and power engineering). Therefore, the situation in the Russian market only reflects the overall trends. However, sooner or later, it will stabilize…” – Oleg Lysak, the director of the additive businesses of the TechnoSpark venture-building company, said.

When the decision was made 2 years ago about investing in the area of additive technologies, it became clear that there was no point in buying the hardware. The least reasonable way to grasp what additive technologies are all about is to purchase a 3D-printer. It is way too expensive and won’t in the slightest help to figure out the market, because once the unit is bought it reduces its practical use exponentially. One should keep a certain market in mind and, moreover, it’s impossible to return the investments made in the purchase of the unit. We decided to follow the way of engineering and joined the efforts with our partner and investor in this group of companies, the Laboratory of computational mechanics (Saint-Petersburg Polytechnic University) headed by Aleksey Borovkov, to establish the CML AT Company specializing in contract-based engineering. What do we mean by that? We have to be second to none when it comes to knowing where, in which cases and how to use additive technologies. That said, it concerns the whole range starting from binder jetting (when printing accessories using sand and polymer resins) all the way to immediate production of finished items of aluminium and titanium. It also includes the use of plastic items, professional and non-professional FDM-printing, SLA – the whole complex of technologies.

In this respect, there is currently a group of developers and product engineers distributed between Moscow and St. Petersburg, and there are more engineers among them who carry out projects on demand from the market, including large companies. For instance, how can we redesign and optimize an item to reduce its production cost? Because in certain cases direct printing doesn’t make any sense. Everyone says: “Let’s print it out!” We respond that it can be done, of course, but the engineering will cost 10 million; printing will require buying a machine that is one of the two machines existing in the whole world; it is very unlikely that you will get any good result and even if you do, the investment will never pay off because you will end up printing 3 items and having troubles with its setting. That is why we suggest using a hybrid approach: say, printing the accessories and optimizing the design so it is fitted for printing…

Sometimes we just say: “Sorry but you have to change the design of the item completely”. And we mean all the surrounding parts, not only some particular supporting piece or another bit; we mean that the whole structure shall be used, all construction shall be altered to produce an absolutely new item and then start manufacturing it by the way of direct metal printing, for example. That is basically what we do. Another important issue is post processing that is often neglected. Despite the fact that the printed items are normally 90% ready, if you operate within the high-tech industry, you will always want to consider their geometry, parameters, dimensional allowances, and you will anyway have to turn to CNC machining. All too often it involves fine or high-precision metal working, which sometimes cancels out all the advantages of 3D-printing in case of an improperly designed item. You have printed items which are worth X amount of money, and metal working costs 5 times more, that all blows up the price of your items to cosmic proportions – and it’s most obviously out of the question.

Slowly but surely, we were figuring out those cases, giving it a go in various markets – medical applications, space, and aviation – but we soon opted out of them because these are very restricted markets that are way too much subject to the geopolitical environment, so to speak. For instance, it is next to impossible for Russian companies to deal with foreign ones in the sector of additive technologies for aviation because it has always been everywhere that the markets of civil and military aviation are really closely… intertwined. So we just went on trying our hands at civil applications: bicycles, railroads, railway machinery, car industry, even iron and steel industry, as strange as it may seem, for instance, we carried out repairs to items for iron and steel industry. The market of repair works, maintenance, spare parts, optimization of all sorts of end bits, tips and nozzles is huge.

Our commercial know-how is a question of criteria; we have developed a sort of matrix of scenarios for the use of different technologies. It serves as a sieve to sift through all of our customers’ items. At first, it is certain expert evaluation: after your technical engineers have seen a piece for a thousand times, they start noticing this and that here and there. Then we very often retrospectively formalize these criteria for future use and further simplification. We are provided with a large number of items – 100, 200, 300 – and we sift them through our sieve. We look at dimensions, materials, permitted admissions, requirements, scenarios of usage, temperatures, capacities of materials, replacement possibilities, criticality of the given part.

The Russian market of additive technologies is virtually nonexistent, there isn’t any at all. So market rules are not applicable. Rather, in most cases they are R&D efforts that are taken in the let’s-try-it-here-let’s-try-it-there fashion. There is only one tiny market segment that really exists, and it is dentistry. Additive technologies haven’t got a big share in the world market either. The whole global market is estimated to be worth 8 billion dollars in spite of all the hype. The Russian market is simply a reflection of the foreign trends, as all the companies view additive technologies as restricted to aviation, space and power engineering. It’s the same all around the world. Take the area of medicine, for example.

The market will sooner or later stabilize as it gets segmented in between those who provide engineering, those who are in charge of the manufacturing process and those who deal with applications. And then at a certain moment it will all be restructured. The situation is the same if you look at the structure of the global market. At first, it was all inside the companies, and prior to that – inside universities. If we look at some statistics, that of the Arcam Company for instance, I assume we will find that it was only in 2015 that the units they sold to the industry outnumbered those they sold to universities. In case of ConceptLaser, the situation manifested itself a bit earlier, in around the year of 2012, when the number of units purchased for scientific purposes diminished by comparison with production. What does it indicate? It just shows that the market hasn’t yet emerged, it’s only now that it is starting to unfold in the true sense of the word.

For example, GE Additive enjoys profits worth $750 million. It is currently one of the largest companies in the world, second only to EOS that boasts $900 million, however GE’s overall profits stand at $118 billion. GE Additive is the department in charge of additive production within the company that is a manufacturer of engines. We take these figures as enormous, though, in fact, the volumes are negligible for the market as a whole. The Russian market hasn’t even made it to this stage of formation so far. Now we are in the phase of quitting the use. There are still people who keep buying the machines to try and figure them out. It seems like they are bought for production, but in reality they serve the let’s-figure-out-if-we-can-print-this-or-if-we-can-print-that purpose… In the meanwhile, the economic side of the matter is left without understanding.

Why did we, for instance, establish TEN.MedPrint? We assume that the market is only going to change once 3D-printing becomes such a straightforward service as, say, CNC-machining. The market bids certain prices for milling operations or lathe works, and these serve everyone as points of reference in other calculations within the production industry, contract manufacturing in particular; we made a conscious decision to establish our company as open contract manufacturing. We haven’t got any captive customers. And our prices are 40% lower than the market average, which anyway satisfies us economically. Ours is not an exponentially but steadily growing trend in terms of profits because we deal with reliable technologies, guaranteed and certified materials, and our customers understand the price formation. In some cases we may not even know who our customers are. We are sent a model, and then reply to their request whether it could be printed or not. If it couldn’t, we say: It needs redesigning!

Contract manufacturing based on 3D-printers is most susceptible to two things. They are payroll expenses and salaries (product designers, salespeople, product engineers, etc. – when a sharp increase in the payroll doesn’t really entail an increase in the output). At the same time additive manufacturing isn’t affected by the price of the powder, it doesn’t depend on it at all. Its share in the production cost of a printed item is 10%. The rest is comprised of the depreciation of the printer, payroll budget and post processing.

The absence of state standards doesn’t critically affect the commercial market. Quite often in case of no other options repair parts are simply used. Then they undergo a set of certain tests that are specified at the beginning, and the tests enable us to use the parts in required products. It means that you print out four items, one of which is tested; in case of a positive outcome each following item is printed together with a control sample. They check the inner structure, modes, quality and give their permission for use. Sometimes customers simply ignore the necessity of certification. Because it may happen to be impossible to produce the item in any other way, or, it may turn out to cost 10 times more. And nobody really cares about certification of bicycles, for example.

GE seems to be the biggest investor in additive technologies in the world. They have invested over $3 billion. Can you imagine one single company investing $3 billion! It is their strategic move, of course, but being proper businessmen, they bought two companies and said that these two companies have to sell 10,000 metal printers over the next 10 years. At that moment both companies were hardly selling more than 150-200 printers in total. It means they announced an exponential growth and added that a good half of it would probably be accounted for by internal demands, by the demands of GE’s partners. The remaining part will require development of markets. It was an entrepreneurial action towards this development.

Speaking of the market of repair works oriented to 3D-printing. I would divide it into 2 sectors. The first one is all about restoration of parts, which basically entails additional printing, and the second one concerns manufacturing of repair and replacement parts. Strangely enough, it involves a great number of consumer goods, for example, repairs to electric scooters… And the first sector is actively growing on the global scale. Take Lufthansa Technik, for example, it was the first company to certify a technology to restore vanes. Not even vanes per se, but the mountings that attach vanes to the bearing ring; it reduced the cost of the vane’s life cycle almost four times. Their useful life is 10,000 hours, then they are extended by smelting, lathed and put back in place – it adds another 10,000 hours. Most obviously, they certified the technology together with the engine producer. It had been preceded by certain friction, though, as Lufthansa Technik is a repair business. Needless to say, producers of new engines are interested in customers changing the engines more often. However, they managed to sort it out through some special authorities and got the readiness of their technology confirmed. Lufthansa has 4-5 printers that are only used for repair works.

If you ask me, the difference between professional-grade and general-use approach is that the professional-grade one ensures reliable quality. Reproducibility of items is 3-4 batches of nine. For plastic items the rate is 2-3 batches of nine, as for metal ones, it stands at 5 batches of nine for this machine and 4 batches of nine for that one. It means that basically one item out of 10 or 10,000 is different from the rest, and that is within quite a tight tolerance interval. The market of plastic printing in Russia is emerging extremely slowly because plastic printing implies general-use items that belong to B2C.

The printer isn’t particularly complex, that is to say, if you have $3 million, you can even construct a 5-axis machine. Within the range of $3 to $10 million, whether you should get a printer or a machine is not the issue. The trouble is to teach it to print, which requires an annual $10 million investment in optimization of printing modes. Elaboration of printing modes implies that the printing processes are tweaked to fit your items in particular. Rate of scanning, laser power, fusion penetration, powder grain size, scanning strategy, number of times a layer is laid by the coating knife, the optimal position of the item on the platform, positioning of the supports, thickness of walls, dimensions of laser focus spot. All in all there are seven or eight parameters (which together comprise the technological engineering process).

Russian engineering centers, though, as well as higher education institutions (Moscow Institute of Steel and Alloys, STANKIN Moscow State Technological University, etc.) are busy teaching printers to print specific items as each one of them has its own operational mode. You have to print out a vane that has a particular cost and meets particular requirements in terms of its porosity, stiffness and so on. One option is what is commonly done here: you can buy a printer and powder, train some product engineers, have a go at printing items this or that way, then five years later eventually understand that nothing works. Three years will cost you money. Another option is to pay, for instance, half a million dollars and get it a month later, as far as I know, you can even choose a printer. I mean you can start producing items for the market half a year later. These are two different strategies.

How is EOS different from Concept Laser? In terms of repeatability. Five versus four item batches of nine. Neither we nor any kind of engineer will ever understand how it is done. It depends only on the workings of the company business. Russian centers specialize in printer design. They may, for instance, have purchased two printers that are collecting dust, and you can’t order anything to be printed out because it is a university. So you say: ‘Let me lease it.’ And they reply: ‘No, you can’t. It is federal property, we will all end up in prison.’ And the printer just won’t print anything. Tomsk Polytechnic University developed an electron beam 3D printer. If Arcam representatives are ever taken there, they will need a lot of cognac to lower the risk of a heart attack, but they will clearly see that there will never be any electron beam 3D printers in Russia. Because they have business and we don’t. It is a matter of investments, a matter of development of printing modes, and making sure the printer is all set to deliver high-quality items on a steady basis. It is a matter of getting the laser settings right. We have a rather vague idea of how deep the debates and talks are when it comes to the geometry of argon flow to the chamber. They modulate the laminar flows to keep the powder in place, laminarity and turbulence to make the unit cool in one place, heat it up in another place and take the temperature in the chamber all at the same time. Can you figure the depth of this method? It means there are no issues that are related to the machine. Their next step is to make the processes fully automatic. It is about handling powders, an automatic chamber, bunkering, loading, unloading and cleaning.

A Dutch startup, the Additive Industries Company, made a modular machine that has only one problem: its capacity is five times that of any other. It is a problem because they can’t provide the needed printing volume for the machine. There is a single loading area followed by two printing modules that have one printing head. Which means while it is doing the printing here, it is making calculations or applying the powder there. This simply makes it work twice as fast. Then the process moves on to simultaneous thermal treating, unloading, cleaning, heat treating and pulling out. In this respect, using this single straightforward process… you get a finished item 5 times faster with fewer people involved. The printing process has been slightly optimized, take the twin laser system, for example; they have also improved the chamber construction and argon flow to ensure the items are of the same quality.

We haven’t experimented with using Russian metal powders in the imported equipment. We are not in the right to do so in terms of business risks. It is better and easier to apply imported powders certified for use in this particular machine, this way we don’t face any risks. Moreover, there are no Russian producers of powders. It is a tiny segment in the whole market of powder materials; there are only three or four producers of good powders for additive technologies in the world. Is China among them? If you are an R&D center, you may give it a go with the Chinese powder. But when you need to get the finished product of a guaranteed quality… For instance, British LPW supplies powders in containers that make it possible to keep track of how the powder is used throughout the whole manufacturing process all the way to the final product. It is essential for the certification process. There are Korean powders and printers, take Insstek, for example. In terms of its design, the machine is R&D-oriented; which makes it a good option for R&D centers but a bad one for batch manufacturing, even if they have managed to sell 40 of those, the units still only suit the research purposes. Because if you look a bit more closely and figure how it works and what kind of components are used to keep it going… It isn’t by chance that it costs this much. I think there are three printer producers; they are EOS, Concept Laser, and SLM (unless they are bought by the Chinese), 3D System can also be named here with a pinch of doubt. All these producers have their own machine strategies too. It takes a simple look at the chambers to see that the machines are oriented towards different markets.

Developing a machine is only the half of it; you also have to make it work steadily. It takes about three to five years to develop such printing modes. It is not that simple. Even GE spends $10-15 million every year to simply develop some specific model software for printing modes, so it was possible to predict what you are going to get (the software is developed by a Russian supplier. They outsource the development of their software to model the printing processes).

The sector of additive technologies has picked a really strange path of development here. There are lots of unfinished R&D efforts that are being taken in total dissociation to what is going on in the industry. In terms of addressing industrial challenges, the level of cooperation with EOS, Fraunhofer Institute or Siemens, for example, is exponentially higher than that with the Russian Academy of Sciences, Sukhoi Super Jet or Aviadvigatel.

The problem with the Russian production industry in general is its vertical integration: since it is impossible to give any economic estimation of the process inside one big holding, it is impossible to set any business goals for 3D printing technologies.

There are five different producers of powders for additive technologies in the world, and there are 10 independent private companies that produce 3D printers. Also, every year we see one or two new startups that emerge to make an attempt at producing printers. That is how our world is made. EOS has such a sale force that it is easier for them to buy the newly emerged startup and integrate it into themselves rather than make a new one; nevertheless, HP managed to come up with a coloured nylon plastic printer that EOS had been trying to make for 20 years. There are now companies like Desktop Metal, Additive Industries or Markforged that specializes in metal printing. The new Metal Alloys startup is what will be slowly but surely bringing forward new applications, as this company is focused on taking applications to new industries. Desktop Metal puts emphasis on other applications, consumer-grade ones, that the EOS company simply can’t embrace because of its size. Bigger companies can’t look out for too many different ideas, it is a normal way of things.

Speaking of the Russian market prospects, I believe that there will be two processes going in parallel. One will be moving along the vertical integration with minimal progress, and at the same time there will be another one developing into other markets with more prospects and a better chance of going global. At a certain moment the processes will merge, because the pendulum pattern of development in our country oscillates between the poles of total engagement and total disengagement. It will either die away, or continue. Right now the Russian market’s growth rate is close to zero. This situation can’t last forever. The market will be expanding over time, we would like it to develop at a quicker pace, of course, but whatever way it goes, the future most definitely lies on the horizontal plane of more specific applications. As soon as we get more horizontal process chains… More steps will be made to create, say, centers for contract printing in robotic machines that will address Russian and global markets on a fifty-fifty basis but will be using Russian materials with highly-qualified Russian technical support. As soon as printing prices get three times lower, and this is now becoming a possibility… I think that promising directions for the export of our contract services in 3D printing will be in the area of medicine, all sorts of customer-made and limited-batch items, and all other repair parts. At the very least, we can expect a breakthrough from the East European market where there are still a lot of Soviet machines. It will only be the beginning, after that it will open possibilities for further development, involvement in larger-scale cooperation.

Sanctions didn’t interfere with our cooperation with GE and Concept Laser. When you are transparent and understandable to them, when you can explain them what you do and how you do it…