Chinese Engine Development

Verum

Junior Member
I totally understand and am agreeing with what everyone is saying. Please let me clear myself. When I mentioned Moore's Law, it's not confined strictly to the idea of doubling every year. I was more just relating to the mere concept. It's no brainer as technology and innovation advance, the acceleration of the rate is itself increasing [the second and third derivative of the rate equation increases].

I totally understand the idea that material research are slow and often yielding with just minimal improvements. But jet engines is the congregation of nearly all modern science. Most improvements come from outside the box. Often improvements come from non-core areas, not just the blade material that we often think of. For example, advances in control system, structural change, fuel technology, and so on. All these could dramatically improve the performance of an engine.

Speaking of semiconductors, I remember a classic case an old Chinese engineer told me before. About 2 - 3 decades ago, China's production yield of silicon ingots used for wafers was always really low, about 20%-30% behind Western counterpart. Doesn't matter what they tried, the problem constantly exists. Later they found out the only thing they needed was a special but simple lid for the furnace. The lid kept the temperature high and produced higher yield. The lack of this simple technique caused Chinese semiconductor industry for at least 10 years.

I'm just sharing a personal view of the matter.
 

broadsword

Brigadier
First off, I'm far from a technical guy unlike most here. But here's my two cents.


However, I have read several well written posts indicating the technological gap in the engine field is widening between Chinese and Western counterparts.

Maybe, you can quote the "well written posts". Provide more details for us to dissect.


Speaking of semiconductors, I remember a classic case an old Chinese engineer told me before. About 2 - 3 decades ago, China's production yield of silicon ingots used for wafers was always really low, about 20%-30% behind Western counterpart. Doesn't matter what they tried, the problem constantly exists. Later they found out the only thing they needed was a special but simple lid for the furnace. The lid kept the temperature high and produced higher yield. The lack of this simple technique caused Chinese semiconductor industry for at least 10 years.

The example you provided above does not do your claim any justice. 2 - 3 decades ago, you can be sure the gap was wider than now. But in your example, you have given us the very reason why the gap has narrowed. Knowledge! Knowledge about the missing lid. 2 - 3 decades ago, how many Chinese had tertiary or polytechnic education? How many went to the West for higher education and came back compared to now? 2 - 3 decades ago, there was no internet in China where they could source for information. China's international trade was a fraction of today's.

But China is not only trading, it is also manufacturing. It is flooding the world with manufactured goods, some of which are high-end. It is a major producer of digital CNC machines, ships, construction machines, etc. It is also making CAT and MRI scanners. China never had those machinery and skill and the high technical education back then to do all these things. I'm sure people who study technology in universities learn about modern technology and maybe research new ones. 2 - 3 decades ago, China did not hold so many patents and the West did not have to cast a wary eye at the rear view mirror for China.

In conclusion why your claim that the gap is widening does not have a leg to stand on and would baffle the learned mind of the West.

But one caveat is that, unlike some people, I believe the gap, even though it is narrowing, will remain after 10 years in most fields. This is because of the sheer number of technical people in the West, the depth and breadth of their knowledge and their high level consumer market. Western companies are always developing their next generation products like Google Glass, whereas China will still be one step behind, at least in high end consumer products. Exception are not many, like the E-cigarettes and electrical pressure cookers.

In a nutshell, the gap is narrowing because the momentum in China is faster than in the West due to its low base.
 

latenlazy

Brigadier
I totally understand and am agreeing with what everyone is saying. Please let me clear myself. When I mentioned Moore's Law, it's not confined strictly to the idea of doubling every year. I was more just relating to the mere concept. It's no brainer as technology and innovation advance, the acceleration of the rate is itself increasing [the second and third derivative of the rate equation increases].

I totally understand the idea that material research are slow and often yielding with just minimal improvements. But jet engines is the congregation of nearly all modern science. Most improvements come from outside the box. Often improvements come from non-core areas, not just the blade material that we often think of. For example, advances in control system, structural change, fuel technology, and so on. All these could dramatically improve the performance of an engine.

Speaking of semiconductors, I remember a classic case an old Chinese engineer told me before. About 2 - 3 decades ago, China's production yield of silicon ingots used for wafers was always really low, about 20%-30% behind Western counterpart. Doesn't matter what they tried, the problem constantly exists. Later they found out the only thing they needed was a special but simple lid for the furnace. The lid kept the temperature high and produced higher yield. The lack of this simple technique caused Chinese semiconductor industry for at least 10 years.

I'm just sharing a personal view of the matter.

If you pay attention to the different performance metrics of each major technology, you will note that not all of them follow the same curve as computing power. Most of those curves actually have a downward slope, not a constant or upward one, which suggests slowdowns in performance gains over time. That all technological innovation has a constant or accelerating pace of improvement is no guarantee. These sorts of things are dependent on the very specific attributes and physics of the technologies involved.
 

siegecrossbow

General
Staff member
Super Moderator
I totally understand and am agreeing with what everyone is saying. Please let me clear myself. When I mentioned Moore's Law, it's not confined strictly to the idea of doubling every year. I was more just relating to the mere concept. It's no brainer as technology and innovation advance, the acceleration of the rate is itself increasing [the second and third derivative of the rate equation increases].

Then maybe you should consider using another example. I think most of the technical guys are just tired of people throwing around terms that they don't completely understand.
 

siegecrossbow

General
Staff member
Super Moderator
If you pay attention to the different performance metrics of each major technology, you will note that not all of them follow the same curve as computing power. Most of those curves actually have a downward slope, not a constant or upward one, which suggests slowdowns in performance gains over time. That all technological innovation has a constant or accelerating pace of improvement is no guarantee. These sorts of things are dependent on the very specific attributes and physics of the technologies involved.

Not to digress too much from topic, but chip design has many trade offs. It isn't always "the faster/more powerful" the better. Should I favor speed but sacrifice reliability and portability due to heating issues? Should I make the chip run faster using completely new design but make life miserable for the programmers because the new instruction set no longer supports legacy code? Should I spend the extra hundred dollars just to improve performance by a slight margin? These are all important questions that engineers need to keep in mind.

That aside, Moore's Law itself may nolonger apply in the near future (barring dramatic technological breakthrough) because you can only shrink the processors down so much before Quantum Effects come into play (there is also heating related issue). Latest processors like Intel's Haswell already require nanometer level precision, and the next generation product will likely hit the limit of what is physically possible. Currently Intel is trying to bypass the problem with 3D lithography, but it is dubious that processing power will increase exponentially for the next few decades.
 

latenlazy

Brigadier
Not to digress too much from topic, but chip design has many trade offs. It isn't always "the faster/more powerful" the better. Should I favor speed but sacrifice reliability and portability due to heating issues? Should I make the chip run faster using completely new design but make life miserable for the programmers because the new instruction set no longer supports legacy code? Should I spend the extra hundred dollars just to improve performance by a slight margin? These are all important questions that engineers need to keep in mind.

That aside, Moore's Law itself may nolonger apply in the near future (barring dramatic technological breakthrough) because you can only shrink the processors down so much before Quantum Effects come into play (there is also heating related issue). Latest processors like Intel's Haswell already require nanometer level precision, and the next generation product will likely hit the limit of what is physically possible. Currently Intel is trying to bypass the problem with 3D lithography, but it is dubious that processing power will increase exponentially for the next few decades.

Oh I'm well aware of the details (though I'm no expert in them). I may not be an EE major but I'm from silicon valley and have been following the silicon industry since high school. We didn't even scratch the surface about all the things that go into what makes your processor faster in this conversation.

I think that they might be able to squeeze out enough innovations to keep Moore's law ticking for another decade, but I agree, we're hitting its limits. That said, if we can make something like graphene work everything changes (not sure just how far off we are from that next ladder step though).
 
Last edited:

Totoro

Major
VIP Professional
Here's a little story that i came up with. It was created by observing the engine deals, chronology of development and so on in Russia and China in the last few decades. While a lot of its content is conjecture, i think there's also a lot of it that sounds quite plausible. Real info is, of course, hard to get. So everyone reading this can make up their minds as to how much of this story is fact, and how much is fiction.

Cold war ended. Saturn, which was developing al31 series of engines as well as producing them with UMPO suddenly found itself without a lot of funding. Salyut, which was just producing al31f, also got itself without much funding.

At the same time, China wanted to modernize its forces. Partly by buying ready products, partly by producing domestic aircraft. Russia was in economic shambles and was desperate for any sort of business. China proposed a few deals which Russia agreed to. One was sale of large amount of al31f, for the su27 that China bought, other was funding of al31fn, a variant of al31f for j10. All that happened in early 1990s.

At the same time, shenyang liming was into full scale development of its ws10 for half a decade. Suddenly, through the al31fn deal, where china financed its whole development, they also bought access to documentation of al-31, possibly even documentation on its production process. At some point, perhaps not right away in 1992 but a bit later, work on a new variant of ws10 started, labeled ws10a. Why shenyang liming stayed with f101 engine core as a basis is unknown, but perhaps too much work was already done with it, and new tech applied from russian producers was seen as enough to finish the ws10 with the core that it already had.

Saturn developed al31fn, and a small preproduction batch of the engine were sent to China to power the j10 prototypes. At the same time, lots of al31f were produced, where both saturn and salyut benefited from it. Saturn used the al31fn money from china to develop not only the said engine, but help develop al31fp for India's Su30. Saturn and Salyut made a deal where Saturn/umpo made engines for india, and Salyut would make engines for china and pay Saturn a provision.

Enter the 2000s. Preproduction of al31fn was done, it was decided salyut would handle the production of serial standard engines. But a spat between saturn and saylut arose. Salyut claimed they had to finish development of al31fn and claimed al31fn as their product. they didn't pay Saturn any royalties from sales. Suddenly there was a big gap in Russian engine producers. Saturn, the big designer and its UMPO cooperant, sponsored by the Russian state, and on the other side the rogue Salyut with relatively scarce resources and very few customers at the time, basically just China.

It is plausible chinese side got a chance for even tighter cooperation with the russian side, but this time directly with Salyut. It provided money and Salyut provided expertise, not only with finishing al31fn and getting it fully ready for j10a, but also with providing assistance with ws10a project.

Of course, such deals are often two way streets. While Saturn was, comparatively speaking, rolling in cash from both Russian state (to develop next gen engine variants) and from lucrative al31fp deals with india and other irkut su30 users, Salyut used the chinese money and work on al31fn variant to further improve al31 family. hence al-31f-m1 finished development around 2006/2007.

Roughly during the same time period, Salyut's work on new al31 variant went into two more projects. One was another subvariant of al31fn, which i'll call "series 2 variant", which was allegedly offered to china in 2006, with 13500 kgf thrust and even thrust vectoring. While we know China didn't go for thrust vectoring addition, we don't know if the engine subariant itself fairly quickly found a way to future j10a built near the end of 2000s and later.

While ws10a allegedly first time flew in 2002., we know it didn't really enter service until the end of that decade. Since Salyut and chinese side enjoyed such good coperation in that period, it is also plausible that not only al31fn work but also work on al31f-m1 actually helped shape ws10a to where it really was once it entered active service on j11b. Anecdotal evidence for that include claims that the engine itself was certified back in 2007., but production quality issues delayed it for a few years. Also, there are claims that russians themselves said ws10a is in the low 130kn thrust class, similar to al31f-m1.

Salyut-Shenyang liming cooperation continued at the end of 2000s and into first half of 2010s. We know bench tests of al31f-m2 were completed in 2012 and flight tests were to commence. Then came news of al-31fn "series 3" engine bench testing in 2013. Evidently, chinese side maintained cooperation and paid for further development all those years. That further makes it more plausible that "series 3" is in fact based on al31f-m2 variant work. official news said "1000 kgf" was added to previous variant, but "series 2" was really never publicized. if news from 2006 about 13500 kgf al31fn variant is true, that plays right into the known 14500 kgf al31f-m2 or hypothetical al31fn series 3 based on it.

At the same time, work on ws10 variants continued. New variant, ws10b, seems to either in production or very near mass production, since a different variant was seen on active duty flankers. if we take into account the fact that ws10a was delayed by quality issues, not the design which was ready some 6 years ago, it is quite plausible another version would be ready in that timeframe. just as salyut went from al31fn to al31f-m1 in 6 or so years, and will take another 7-8 years to m2 variant. While ws10b is likely to contain more domestic input now than russian, and perhaps is not quite up to m2 variant level, it may well be in 135-140 kn range. Since it was not known how that'd turn out, years earlier the decision to pay salyut for al31fn series 3 was made. now the chinese side may find itself in a situation where they have two engines of comparable perfomance to choose from.

Also, while ws15 may be based on a different core, perhaps domestic one, it too is bound to have benefited from all this cooperation. But who knows when it will enter mass production.

Some of this fiction was sourced from here, a page aggreggating some other sources on al31 history. Other stuff was concocted in my head.

Please, Log in or Register to view URLs content!
 

latenlazy

Brigadier
Here's a little story that i came up with. It was created by observing the engine deals, chronology of development and so on in Russia and China in the last few decades. While a lot of its content is conjecture, i think there's also a lot of it that sounds quite plausible. Real info is, of course, hard to get. So everyone reading this can make up their minds as to how much of this story is fact, and how much is fiction.

Cold war ended. Saturn, which was developing al31 series of engines as well as producing them with UMPO suddenly found itself without a lot of funding. Salyut, which was just producing al31f, also got itself without much funding.

At the same time, China wanted to modernize its forces. Partly by buying ready products, partly by producing domestic aircraft. Russia was in economic shambles and was desperate for any sort of business. China proposed a few deals which Russia agreed to. One was sale of large amount of al31f, for the su27 that China bought, other was funding of al31fn, a variant of al31f for j10. All that happened in early 1990s.

At the same time, shenyang liming was into full scale development of its ws10 for half a decade. Suddenly, through the al31fn deal, where china financed its whole development, they also bought access to documentation of al-31, possibly even documentation on its production process. At some point, perhaps not right away in 1992 but a bit later, work on a new variant of ws10 started, labeled ws10a. Why shenyang liming stayed with f101 engine core as a basis is unknown, but perhaps too much work was already done with it, and new tech applied from russian producers was seen as enough to finish the ws10 with the core that it already had.

Saturn developed al31fn, and a small preproduction batch of the engine were sent to China to power the j10 prototypes. At the same time, lots of al31f were produced, where both saturn and salyut benefited from it. Saturn used the al31fn money from china to develop not only the said engine, but help develop al31fp for India's Su30. Saturn and Salyut made a deal where Saturn/umpo made engines for india, and Salyut would make engines for china and pay Saturn a provision.

Enter the 2000s. Preproduction of al31fn was done, it was decided salyut would handle the production of serial standard engines. But a spat between saturn and saylut arose. Salyut claimed they had to finish development of al31fn and claimed al31fn as their product. they didn't pay Saturn any royalties from sales. Suddenly there was a big gap in Russian engine producers. Saturn, the big designer and its UMPO cooperant, sponsored by the Russian state, and on the other side the rogue Salyut with relatively scarce resources and very few customers at the time, basically just China.

It is plausible chinese side got a chance for even tighter cooperation with the russian side, but this time directly with Salyut. It provided money and Salyut provided expertise, not only with finishing al31fn and getting it fully ready for j10a, but also with providing assistance with ws10a project.

Of course, such deals are often two way streets. While Saturn was, comparatively speaking, rolling in cash from both Russian state (to develop next gen engine variants) and from lucrative al31fp deals with india and other irkut su30 users, Salyut used the chinese money and work on al31fn variant to further improve al31 family. hence al-31f-m1 finished development around 2006/2007.

Roughly during the same time period, Salyut's work on new al31 variant went into two more projects. One was another subvariant of al31fn, which i'll call "series 2 variant", which was allegedly offered to china in 2006, with 13500 kgf thrust and even thrust vectoring. While we know China didn't go for thrust vectoring addition, we don't know if the engine subariant itself fairly quickly found a way to future j10a built near the end of 2000s and later.

While ws10a allegedly first time flew in 2002., we know it didn't really enter service until the end of that decade. Since Salyut and chinese side enjoyed such good coperation in that period, it is also plausible that not only al31fn work but also work on al31f-m1 actually helped shape ws10a to where it really was once it entered active service on j11b. Anecdotal evidence for that include claims that the engine itself was certified back in 2007., but production quality issues delayed it for a few years. Also, there are claims that russians themselves said ws10a is in the low 130kn thrust class, similar to al31f-m1.

Salyut-Shenyang liming cooperation continued at the end of 2000s and into first half of 2010s. We know bench tests of al31f-m2 were completed in 2012 and flight tests were to commence. Then came news of al-31fn "series 3" engine bench testing in 2013. Evidently, chinese side maintained cooperation and paid for further development all those years. That further makes it more plausible that "series 3" is in fact based on al31f-m2 variant work. official news said "1000 kgf" was added to previous variant, but "series 2" was really never publicized. if news from 2006 about 13500 kgf al31fn variant is true, that plays right into the known 14500 kgf al31f-m2 or hypothetical al31fn series 3 based on it.

At the same time, work on ws10 variants continued. New variant, ws10b, seems to either in production or very near mass production, since a different variant was seen on active duty flankers. if we take into account the fact that ws10a was delayed by quality issues, not the design which was ready some 6 years ago, it is quite plausible another version would be ready in that timeframe. just as salyut went from al31fn to al31f-m1 in 6 or so years, and will take another 7-8 years to m2 variant. While ws10b is likely to contain more domestic input now than russian, and perhaps is not quite up to m2 variant level, it may well be in 135-140 kn range. Since it was not known how that'd turn out, years earlier the decision to pay salyut for al31fn series 3 was made. now the chinese side may find itself in a situation where they have two engines of comparable perfomance to choose from.

Also, while ws15 may be based on a different core, perhaps domestic one, it too is bound to have benefited from all this cooperation. But who knows when it will enter mass production.

Some of this fiction was sourced from here, a page aggreggating some other sources on al31 history. Other stuff was concocted in my head.

Please, Log in or Register to view URLs content!
Some parts of this story don't jive with some of the facts about the WS-10s development, including that the initial problem for the WS-10 was a slow spool time (I believe this was 2006) and the WS-10A was in part meant to rectify that problem.
 

Totoro

Major
VIP Professional
As far as i know, spool up time is still an issue with ws10a, thats why it isnt used on single Engined planes nor carrier planes. It may be rectified with ws10b, we will see.

Ws10a flew around 2002 and its development ended around 2007, butproduction quality issues werent resooved until 2009 or so. So that part matches the story.
 

Blitzo

Lieutenant General
Staff member
Super Moderator
Registered Member
As far as i know, spool up time is still an issue with ws10a, thats why it isnt used on single Engined planes nor carrier planes. It may be rectified with ws10b, we will see.

Care to elaborate on that logic?

I.e.: why would longer supposed spool time be an issue for carrier based fighters and single engine fighters but not land based twin engine fighters?
 
Top