Still waiting..

[pmplayer 9,589]

..for a update from Fenix 🤔

Actually less because of the general code, which is actually fine. But the second engine variant (IAE) would now be desirable.

Thought the IAE Variant would come sooner, but it's been going on for a long time now..

What do you guys think about it ?

 

cheers 😉

 

 

[SierraDelta 1,352]

6 minutes ago, pmplayer said:

What do you guys think about it ?

I think you should just check this now and then:

https://discord.com/channels/736572376967282769/736710909887643679

[G-YMML1 584]

Remind me please, how long did it take for FSLabs to bring sharklet version for their A320? I think it's been more than 2 years. Then 319 and 320 followed quite fast? 

[polosim 858]

From Aamir:

… “

I’ll start with the not-so-great news, progress on the IAEs (and by extension - the external engine model) has been slower than expected due to the conflict in Ukraine. The developer taking lead on the integration of the external engine model is, unfortunately, regularly without power and heating for days. The recent large scale missile strikes in the region have only served to worsen the situation, and we do not expect the circumstances to improve in the near future. With that being said, work does continue where possible, and last night we managed to clear an important prototype milestone - the implementation of fast lookup algorithms, making the engine model math more performant by nearly 1000x.

I have not shared many details about the engine model as yet, but we are being ambitious with our goals - and the set of data we’re using to build it is absolutely enormous. This fast lookup algorithm clears a large hurdle in making the implementation possible. We are inching closer to having a prototype flying, but as ever, we will not be providing a timescale for this given the variable circumstances around it.

Please take this as a subtle reminder that while we are all extremely committed to what we do here at Fenix, sometimes life takes a front seat instead, and the welfare, comfort, and safety of our developers will always come first.” …..

[pmplayer 9,589]

38 minutes ago, polosim said:

From Aamir:

… “

I’ll start with the not-so-great news, progress on the IAEs (and by extension - the external engine model) has been slower than expected due to the conflict in Ukraine. The developer taking lead on the integration of the external engine model is, unfortunately, regularly without power and heating for days. The recent large scale missile strikes in the region have only served to worsen the situation, and we do not expect the circumstances to improve in the near future. With that being said, work does continue where possible, and last night we managed to clear an important prototype milestone - the implementation of fast lookup algorithms, making the engine model math more performant by nearly 1000x.

I have not shared many details about the engine model as yet, but we are being ambitious with our goals - and the set of data we’re using to build it is absolutely enormous. This fast lookup algorithm clears a large hurdle in making the implementation possible. We are inching closer to having a prototype flying, but as ever, we will not be providing a timescale for this given the variable circumstances around it.

Please take this as a subtle reminder that while we are all extremely committed to what we do here at Fenix, sometimes life takes a front seat instead, and the welfare, comfort, and safety of our developers will always come first.” …..

OK, did not tought that, there realy a lot Developer from Ukraine around the Dev. Teams.

Anyway thanks for reply - so we have to wait.. ( By the way, what a stupid and ridiculous war, unbelievable !!! )

 

cheers 😏

[Noel 4,554]

19 minutes ago, pmplayer said:

By the way, what a stupid and ridiculous war, unbelievable

Uh oh, you're commenting on politics here in fantasyland let's see how long it takes before....

[Aamir 2,907]

Early into the project I had a decision to make, which was namely - do the IAEs quickly or do the IAEs properly. There wasn't an option to go in-between - I mean, we could have thrown it into the MSFS FDE and hammered it into something "acceptable", in that it looks like an IAE, the values would be close-ish, but it would more or less stop there. There would not be a lot more depth, the startups would run on a script to emulate the rough character of an IAE start, we'd script in some timings for spool etc to make it all tick over and call it a day, fuel flow etc would be okay-ish, much like the CFMs, perhaps within 10% of a margin. After all, most had already purchased the product and commercially it makes little sense to go out and try and push the boundary because.. well, frankly we weren't going to get paid for it. So doing the bare minimum was probably the most efficient, commercially rewarding option. With that being said I also recognised that people pay us to go out and find that last tenth. I daresay that's what our customers would expect. So I made the call to go for the hard option. Do it properly. As much as being the director of the company forced me into considering the commercial impact of this, there's something to be said for taking appropriate care and responsibility in maturing a product that, ultimately, customers are invested in. Maturing a product in software takes time, it doesn't matter how many resources you throw at it. An experienced project manager will probably tell you adding too many resources and just throwing bodies at the problem is likely to exacerbate the problem more than speed anything up, too. We've not yet even crested the 1st anniversary of our very first product being out in the market. PMDG, for example, launched their first iteration of 737 in 2003, they'll be celebrating 20 years of that thing being out there in some form or the other, before we celebrate our first - as an example. That's a lot of time to refine, renew, rewrite, etc - and as you can see it's a superb, mature product for it. 

That latest post that @polosim quoted was from a couple of months ago, and since then we've prototyped the engine model in it's steady state, it's working fine and flying around, we're reasonably happy with the fuel flow and what not. That's the easy part. The real challenge lies with modelling transient states of a turbofan. For example, programming a startup is less script and more physics based. For example a thermal model for EGT cooldown and rise during shutdown and startup is crucial given the IAEs motor the engine if EGT > 250 degrees when start is commanded, so you need to model the air mass effects of the motoring and how the air driven by the fan blade spinning cools the engine, to add to that, you need to accurately place the EGT probe for the actual sensor as the sensor itself isn't in the engine core, so if you stop motoring the engine the temperature fluctuates accordingly as heat transfer changes depending on airflow. If you dry crank a hot engine (right after shutdown), due to where the EGT sensor is (hint: the exhaust), the core will actually blow hot air from the core outwards and backwards over the EGT probes so you end up cranking the engine and the EGT rising for a very short period of time, before cooling. Those EGT probes are over by the exhaust strut, quite far from the core, so on shutdown the EGT reading actually shows it being quite fast to cool as a result, too, however not before slightly rising as the N1 fan and N2 rotor stop spinning and thus providing some airflow over the probes, so there's a little temporary spike. You cannot get these details right on a simple simulation.

This is before you start thinking about starting or shutting down an engine mid flight, as the fan windmills at 24,000ft. A whole different ballgame there, but there are a number of startup modes in that case, assisted starts, unassisted starts, depending on the N1, which depends on speed, alt, pressure, etc, which then in turn affects how the engine behaves for all parameters on a start.

Even down to the oil. IAE engines do this signature oil "gulp" when they start. Oil goes down then back up during a start. We know that, but why, and how much? We can't just fake up data and call it a day. We need to understand why and model it's effects otherwise the model will fall apart. This effect comes into place because the oil pump is attached to a gearbox which is N2 driven, as it all starts turning on a startup the quantity of the oil tank starts decreasing for a little while because the scavenge flow from the bearings must travel through several pipes and then back into the tank. Related to N2, so the effects on N2 on a start will affect this also.

All of this without modelling the pesky EEC and things like the IAE's Keep Out Zones, where the FADEC will flat out refuse to stablise the engine at between 60 and 74% N1 due to fan flutter, which is a whole other kettle of fish. 

All of this data, absolutely all of it, we've gathered on the real deal. A level-D doesn't give us this much insight and data into what/why/when/how. We needed the real engine to show us what we needed to know. This goes beyond most pilot's comprehension of their equipment, so it's not like it's knocking around in anyone's head either. This is a tiny, small slice of insight into what building a model like this entails. There's dozens upon dozens of little snippets like this that make up the entire thing, and it's safe to say building it is an incredible challenge, even just from a data acquisition, management and organisation front, forget about the actual legwork of coding the thing after the fact. 

I hope from the little extract above, you can gather why this is taking months of work. But when it happens, it'll be worth it. We wanted to build it because it's frankly just flippin' cool. Forget all the commercial implications, at the end of the day it's satisfying building something you can sit back in your chair and go: "Dang..", after all, we're enthusiasts all said and done too..  

[Mike S KPDX 598]

I dont want to be combative at all but just very curious.  Why the big push for this engine type?  I wonder what significant change one would see within the cockpit or general operation.  A 321 or A319, sure, I get it... but engine type??

[Aamir 2,907]

Just now, Mike S KPDX said:

I dont want to be combative at all but just very curious.  Why the big push for this engine type?  I wonder what significant change one would see within the cockpit or general operation.  A 321 or A319, sure, I get it... but engine type??

Two for one - we can build a framework which can apply backwards to the CFMs and we can bring those up to a really superb standard also. 

A321 and A319 have different engines with minutely different characteristics, but this will allow us to dive into those at a higher level also. 

[lwt1971 7,455]

9 minutes ago, Aamir said:

That latest post that @polosim quoted was from a couple of months ago, and since then we've prototyped the engine model in it's steady state, it's working fine and flying around, we're reasonably happy with the fuel flow and what not. That's the easy part. The real challenge lies with modelling transient states of a turbofan. For example, programming a startup is less script and more physics based. For example a thermal model for EGT cooldown and rise during shutdown and startup is crucial given the IAEs motor the engine if EGT > 250 degrees when start is commanded, so you need to model the air mass effects of the motoring and how the air driven by the fan blade spinning cools the engine, to add to that, you need to accurately place the EGT probe for the actual sensor as the sensor itself isn't in the engine core, so if you stop motoring the engine the temperature fluctuates accordingly as heat transfer changes depending on airflow. If you dry crank a hot engine (right after shutdown), due to where the EGT sensor is (hint: the exhaust), the core will actually blow hot air from the core outwards and backwards over the EGT probes so you end up cranking the engine and the EGT rising for a very short period of time, before cooling. Those EGT probes are over by the exhaust strut, quite far from the core, so on shutdown the EGT reading actually shows it being quite fast to cool as a result, too, however not before slightly rising as the N1 fan and N2 rotor stop spinning and thus providing some airflow over the probes, so there's a little temporary spike. You cannot get these details right on a simple simulation.

This is before you start thinking about starting or shutting down an engine mid flight, as the fan windmills at 24,000ft. A whole different ballgame there, but there are a number of startup modes in that case, assisted starts, unassisted starts, depending on the N1, which depends on speed, alt, pressure, etc, which then in turn affects how the engine behaves for all parameters on a start.

Even down to the oil. IAE engines do this signature oil "gulp" when they start. Oil goes down then back up during a start. We know that, but why, and how much? We can't just fake up data and call it a day. We need to understand why and model it's effects otherwise the model will fall apart. This effect comes into place because the oil pump is attached to a gearbox which is N2 driven, as it all starts turning on a startup the quantity of the oil tank starts decreasing for a little while because the scavenge flow from the bearings must travel through several pipes and then back into the tank. Related to N2, so the effects on N2 on a start will affect this also.

All of this without modelling the pesky EEC and things like the IAE's Keep Out Zones, where the FADEC will flat out refuse to stablise the engine at between 60 and 74% N1 due to fan flutter, which is a whole other kettle of fish. 

All of this data, absolutely all of it, we've gathered on the real deal. A level-D doesn't give us this much insight and data into what/why/when/how. We needed the real engine to show us what we needed to know. This goes beyond most pilot's comprehension of their equipment, so it's not like it's knocking around in anyone's head either. This is a tiny, small slice of insight into what building a model like this entails. There's dozens upon dozens of little snippets like this that make up the entire thing, and it's safe to say building it is an incredible challenge, even just from a data acquisition, management and organisation front, forget about the actual legwork of coding the thing after the fact. 

I hope from the little extract above, you can gather why this is taking months of work. But when it happens, it'll be worth it. We wanted to build it because it's frankly just flippin' cool. Forget all the commercial implications, at the end of the day it's satisfying building something you can sit back in your chair and go: "Dang..", after all, we're enthusiasts all said and done too..  

Absolutely take all the time you need to simulate and model engine behaviour/physics to the level of such impressive detail you've described above! Certainly is in keeping with the theme and depth of simulation the Fenix is all about. "Dang.." is an apt term 🙂, Holy **** is another.
 

[polosim 858]

18 minutes ago, Aamir said:

Early into the project I had a decision to make, which was namely - do the IAEs quickly or do the IAEs properly. There wasn't an option to go in-between - I mean, we could have thrown it into the MSFS FDE and hammered it into something "acceptable", in that it looks like an IAE, the values would be close-ish, but it would more or less stop there. There would not be a lot more depth, the startups would run on a script to emulate the rough character of an IAE start, we'd script in some timings for spool etc to make it all tick over and call it a day, fuel flow etc would be okay-ish, much like the CFMs, perhaps within 10% of a margin.

It is nice to know all the effort and realism that is being considered in the development of the IAE engine. On the other hand, what progress has been made in improvements with respect to the CFM engine, for example its poor performance in single-engine operations.
Wishing you all the success you deserve in the development of your product.

[Flightsky 77]

21 minutes ago, Aamir said:

Two for one - we can build a framework which can apply backwards to the CFMs and we can bring those up to a really superb standard also. 

A321 and A319 have different engines with minutely different characteristics, but this will allow us to dive into those at a higher level also. 

"Delve into those at a higher level also."

So the A321 and A319 are projects down the road for the Fenix team, am I right 😁

[Aamir 2,907]

1 hour ago, polosim said:

It is nice to know all the effort and realism that is being considered in the development of the IAE engine. On the other hand, what progress has been made in improvements with respect to the CFM engine, for example its poor performance in single-engine operations.
Wishing you all the success you deserve in the development of your product.

Progress on the IAEs is progress on the CFMs. Why build all this infrastructure only to not use it for CFMs also? They will go through the same extensive update, but without the time sink of building the framework and integrations after we're done with the IAEs.

[FrankR409 514]

2 hours ago, Mike S KPDX said:

I dont want to be combative at all but just very curious.  Why the big push for this engine type?  I wonder what significant change one would see within the cockpit or general operation.  A 321 or A319, sure, I get it... but engine type??

You have to admit, the gulping oil sound at startup, worth it for me and I don’t even own the plane (yet)