Ki-84 WEP Total Duration

[nrshida]

Referring to Lusche’s latest WEP chart (http://bbs.hitechcreations.com/smf/index.php/topic,386482.0.html) - based on HiTech’s previous data (http://bbs.hitechcreations.com/smf/index.php/topic,383002.msg5097338.html#msg5097338). I noticed an inconsistency which I suspect must an error in the ‘total time’ of WEP. There is a pretty obvious inconsistency with some of the other aircraft. Well, three to be specific.

The Ki-84's WEP was achieved in real life through water / methanol injection. Many of the German V12s used a very similar system. Let’s just call that MW for convenience (I know the exact constitution varied subtly but it's not relevant).


A quick comparison then of MW tank capacities and total AH WEP durations from three examples:


Bf109K-4 - 43 minute total WEP time. MW-50 tank capacity - 118 litres (31.2 US liquid Gallons).

Fw190D-9 - 43 minute total WEP time. MW-50 tank capacity - 115 litre capacity (30.37 US liquid Gallons).

The American Technical Intelligence report (T-2 Report on Frank-1 (KI-84), T-2, Serial MO. 302, Interim Report NO. 3 (Project NO. NAD 25 - Release Date November 1946) Page 3 - 'factual data' states the water and methanol tank of the Ki-84 had the capacity of 42.2 gal. Which is 159.74 litres. Total WEP time in AH: 10 minutes.


To recap:


Bf109K-4 - 43 minute total WEP on 118 litres (31.2 US liquid Gallons) of MW.

Fw190D-9 - 43 minute total WEP on 115 litres (30.37 US liquid Gallons) of MW.

Ki-84 - 10 minute total WEP on 160 litres (42.4 US liquid Gallons) of MW.


The displacements of the engines are comparable:-

Daimler-Benz DB 605 - 35.7 litres.
Jumo 213E - 35 litres.
Nakajima Homare 21 - 35.8 litres.

All three engines used direct fuel injection. The German engines are 12 cylinder while the Nakajima an 18. Additionally the Nakajima 2-valve head, German V-12s 4 valve heads. All were supercharged. The Nakajima was air / oil-cooled, German ones liquid-cooled. The Ki-84 is heavier than the 109 due to a heftier structural construction, slightly larger size and the hydraulic system - don’t know about the 190s.

Even conceding some drop in efficiency to allow for the above factors the Nakajima has 36% and 39% more supply of WEP juice than the Fw190 and Bf109 respectively. Yet the new total WEP scheduling suggests the following relative rates:-


43 minutes x 60 = 2580 seconds of WEP total for the Bf109K-4 - That’s 2.7 litres per minute consumption of MW alone (fuel & air aside).

10 minutes x 60 = 600 seconds of WEP total for the Ki-84 - That’s 16.0 litres per minute of MW alone (fuel & air aside).


I can’t give actual WEP consumption rates because that information is lost but making inferences from the K-4 seems sound given how similar in execution it was and how well documented the German engineering was / is. Intuitively, just considering cylinder displacement, revolutions, power cycles and fuel mixtures from other engines it looks likely the Homare couldn’t possibly consume petrol (avgas) at that rate, let along petrol PLUS the MW. At the consumption rate AH now models that is 4 (US) gallons per minute, of liquid additive alone!


I am aware there is a lot of rumbling in the AH jungle about the WEP modelling of many aircraft. I don’t know enough detail about other types to comment. Obviously a balance has to be struck for aircraft in which pilots were instructed to only go to a point for so long to preserve engines and some throttles had breakable wires etc. In AH we’d likely be there all the time. If however you are going to model the consumption of a finite medium, then surely it cannot be done with any semblance of realism without carefully considering the capacities and consumption rates.

I hope this post therefore makes a convincing argument that the total WEP time for the Ki-84 ought to be corrected to somewhere considerably closer to the K-4 end of the spectrum.

[BuckShot]

Doesn't this assume that the german and Japanese wep juice are drank at the same rate?

Maybe the ki had more thirst per output.

[Copprhed]

Shida knows his Ki 84s.

[nrshida]

Doesn't this assume that the german and Japanese wep juice are drank at the same rate?

Maybe the ki had more thirst per output.

Oh great Scott

[nrshida]

I have a theory where the notion of a small WEP supply might originate. Some of the best pictures inside the aft fuselage are through the radio hatch of the only remaining Ki-84 when it was in flying condition (in the 80s). That small alloy cylindrical tank which is the only visible tank in the same approximate location as the MW tank is the hydraulic fluid reserve reservoir and NOT the MW tank. The latter would have been mounted between frame 8 and 9 and not 9 and 10 - 10 being the front limit of the radio access hatch. Frames 1, 2, 5 and 9 are the significant structural points in that segment of the fuselage. Additionally the MW tank extends all the way up to shoulder level. It had been removed from that aircraft by that point.

If you estimated the capacity of that hydraulic tank (which could be done proportionately to the prototype drop tank, say), assuming that it was the MW tank, that might give you a small capacity / duration in keeping with 10 minutes of total WEP.

[nrshida]

Further research reveals more AH WEP modelling weirdness…

I read only the Model III Ki-43 ‘Oscar’ incorporated a water / methanol injection system so the WEP in the AH Ki-43-II must be of the ‘engine over stressing’ kind. Similarly I can’t find a MW container at all in the Bf109-E, there are compressed air bottles in the expected area. For the MG-17s apparently. I think it's unwise to get discussions of the different kinds of WEPs mixed up. Here I'm primarily interested in the finite consumable-based WEP systerms, since placing an upper limit on these is a recent initiative by HTC. I'm assuming it's a step towards greater realism.

I looked into the N1K2 and found a MW tank capacity listed at 140 litres. This was even brought up on this very forum in 2002 (http://bbs.hitechcreations.com/smf/index.php/topic,53815.0.html). The N1K2 uses exactly the same engine as the Ki-84 so while a different total time could be attributed to a slight difference in capacity of MW, the vastly different use and cooling cycles are a mystery. Doubly so because the cooling system is apparently identical (N1K experts please correct me if I'm wrong).

AH N1K2 - 5 minutes WEP, 17 minutes cool down and 10 minute total.
AH Ki-84 - 2 minutes WEP, 1 minute cool down and 10 minute total.


The War Thunder developers also had to make choices regarding how they model WEP. Some of my searches drew me to their forum today - it seems they also tend to use historical documents and descriptions to make these choices. Some of the developers seemed to join into some of the discussions, especially the K-4 ones. Over there the Ki-84 ko (basically what we have) gets 36 minutes of WEP and the N1K2-J 35 minutes.

From this I infer they’ve modelled the finite MW capacities with reference to the capacities of the real aircraft.

N1K2 - 35 minutes (2100 seconds) on 140 litres - 0.067 per second, 4.02 litres per minute.
Ki-84 - 36.76 minutes (2205 seconds) on 160 litres - 0.072 per second, 4.32 litres per minute.

Interestingly they model the Bf109K-4’s WEP at a total of 21 minutes.

K-4 - 21 minutes (1260 seconds) on 118 litres - 0.094 per second, 5.64 litres per minute.


Like I said earlier due to certain factors the 109’s volumetric efficiency might be higher allowing a greater flow rate of charge cooling. However these consumption rates and capacities do corelate fairly well and reenforce the absurdity of a 10 minute total time for the Homare-engined aircraft.


While I’m by no means saying that War Thunder is the epitome of correct modelling, those total times seem more in keeping with the size of the MW tanks and design philosophies. These figures are close to the calculations (including some estimates) I made before I even looked there.



Respectfully then I ask that the total WEP times which now model consumable-based WEP be re-examined in a less arbitrary way than they presently appear to be.

[USCH]

  take note guys, this is how you make a case.

[nrshida]

take note guys, this is how you make a case.

Thank you sir, it's a lot of work and I feel like one of those nutters standing on a box in Hyde park, like no one is listening! But bloody-minded is as bloody-minded does so let's press on...

What this thread needs is pictures!

[nrshida]

So this has been really bugging me. I’ve searched for days to find an original picture of the water / methanol tank. I know it exists in at least one book - but I can’t obtain a copy. I think I may have even have had it once but lost it along with my old computer and an attack of appathy.

This is rather convoluted but I hope it helps convince…

I found this Greek model maker who has God-level scratch-building skills. In a thread dated 2015 he documents a build he made of a 1:48th scale Hasegawa Ki-84, lots of panels off to show the interior detail:-






Have you ever seen such dedication and attention to detail?! His name is Panagiotis from Athens. These pictures are his property.

You can follow the build here:

https://ww2aircraft.net/forum/threads/nakajima-ki-84-hayate-frank-1-48-hasegawa.43836/page-3

Unfortunately the annotated picture he posted asking what the two tanks were in this section of the fuselage is now a broken link. Someone responded: "Methanol Tank(left circle) and Hydraulic Oil Tank(right circle)". Panagiotis scratch built them as they aren't in the kit.







The MW tank is the blueish-green, lacquered, larger cylinder on the left, just overlapping the cockpit floor, just in front of the radio access hatch. What I said earlier suggesting the MW tank extended up to shoulder level was based on a poor-quality exploded view - all I had - and is clearly erroneous. It does however sit between frames 9 and 10. Metal straps transfer the load as it would have been quite heavy when full.

I wanted to try and calculate the volume of the tank. To my eye even that didn’t look large enough to hold 160 litres. What I did was to take the Verniers to my own Hasagawa 1:48th Ki-84 model (still under construction) and estimate dimensions based on parts I do know the dimensions of like the seat for example.

On my first attempt I came up 20 litres short of 160 litres. I was surprised thinking that 160 litres would need an enormous tank. Not so. Seems my muddled brain thinks in Imperial gallons for petrol tanks and only litres for coke bottles. I also realised the Bf109 fuselage is a lot smaller than the Ki-84. Makes the tank look bigger in that aircraft. There’s never any frame of reference in these picture hunts.

‘Walking-to-target’ I found that a diameter of 320-mm x height 500-mm gave 141428571.43 cubic mm or 160.91 litres. Dimensions like these look reasonable inside the Hayate’s fuselage. Incidentally I also found a British report which suggested a capacity of 35 gallons. When I compared US and Imperial gallons the capacity still corelated quite well: 159 litres. Probably small rounding issues explain the 1 litre difference.

[nrshida]

So compelled by a niggling feeling in the subconscious for several days I had to do some further research and now have to correct some of my ‘off-piste’ detective work I did for the last post. Trying to do this on the fly with slim to no historical information, so please bear with me.

My niggling thought since I posted it: Where's the filler cap for the MW tank? Why would you necessitate opening the radio hatch and pouring a water-methanol mix with a funnel or can in that awkward space? 160 litres of MW would weigh somewhere around 120 kgs. There’s just no way you’d be filling that from a single container in that space. And in a hurry you’d be spilling it all over the radio and CD player. Also, MW injection was in the Ki-84 right from the start (Ki-43 only from Model III onwards). Why wouldn’t you add an external filler cap?

Then there are these two pictures of a Dora and G-10 respectively:






The tanks look proportionally bigger than the one in Panagiotis’ model yet in real life they are smaller.

So I rechecked my sums. 320 x 500 gives 40212385.97 cubic millimetres not 141428571.43. That’s eight significant figures not nine and only 40 litres.

Would need a cylinder with something like 550-mm diameter and a length of 677-mm to clear 160 litres. That lacquered tank is nowhere near that big though. I must have made a mistake with using the online calculator for volume of cylinders - perhaps I was putting diameters into the radius box - that's why the output looked reasonable.

Anyway since the capacity of 160 litres is given in reputable documents there can only be one conclusion - that isn’t the water-methanol tank.

This is the diagram I referred to earlier:



Yes I know. Like trying to play 'Where’s Wally / Waldo' isn’t it. The item we’re interested in is labelled 60. I zoomed in for those who are interested:-



Hard to make out any detail at all. Basically one small line behind frames. Can't even make out the shape. Might be a cylinder, more likely a rounded box-shape. This cutaway view doesn't have Panagiotis' tank detailed either, and I've seen that in other real aircraft - exactly like that, just different colours.

So unfortunately unless anyone else out there has some historical pictures I don’t have access to I cannot provide a visual reference of the Ki-84 water-methanol tank. The last existent Ki-84 doesn’t have one anymore. Not for over 35 years at least. Some dingleberry’s probably brewing moonshine in it to this day not knowing what it is. Or more likely it split in the late 1940s and got thrown in a dumpster.

Could also explain the vagueness and contradiction in the texts. Some place the MW infront of the armour. The last Ki-84 doesn't have its armour either


Doesn’t affect the case for revised WEP. I was just going the extra mile in the last post to show it visually. Have to honestly correct my own mistakes or it’s straight to cartoon hell with the vulchers and gangtards when I die. Here is the original document which lists the capacity:

http://www.mediafire.com/file/hikoc0d754qcf94/1946_AAF_T-2_FRANK_KI-84_INTERIM_REPORT_%233--SECURE.pdf

Top of page 3 - I think this has to be public domain by now. If not come and track me down and we'll fight about it. I must be easier to find than a Ki-84 MW tank   

[hotcoffe]

have anyone noticed the actual KI-84 really looks like a 190 but in game they look totally different ?

I wonder if it was based on 190 designs supplied to Japs by Germans or vice versa was the 190 based on KI designs supplied to germans by japs ?

[nrshida]

By courtesy of Gman I was able to find the following information from Jane’s Fighting Aircraft of World War II - Bill Gunston (ISBN: 9780517679647). About the German MW system:-


"The system was used to obtain extra power below the rated altitude of the engine.  The mixture was injected into the intake side of the supercharger and acted as an anti-detonant, providing charge cooling and enabling higher boost pressures to be used”.

Basically an intercooler substitute then.


“The increased power could be used for a maximum of 10 minutes and at least 5 minutes had to elapse between successive periods of operation”.

"On the Me 109, injection into the supercharger of the DB 605 MW was at the rate of approximately 35 gallons per hour. The normal fuel consumption at the take-off rating was 106 gallons per hour, but this was increased to 141 gallons per hour when using the MW 50 system with higher boost pressure”.

"The Ta 152 B aircraft with a DB 603 E engine had a maximum speed of 417 m.p.h. at 27,000 ft.  When the GM 1 system* was used, at an injection rate of 32.2 lb./min. the maximum speed was increased to 348 m.p.h.”.


 - All from Page 288


* Please note the GM1 system was NOT an MW injection but a Nitrous-oxide injection system (code-name “ha-ha”). This was also injected into the supercharger but remained in liquid form under pressure. I don’t know which Ta 152 AH models but the one described apparently had a different WEP system from the Bf109.



Now addressing the rate of consumption:-

just have to be careful here because Bill Gunston was English and likely those figures are Imperial Gallons. Let's keep it all in US liquid Gallons:-

35 Imperial Gallons = 42.0332 US liquid Gallons let's call is 42. 42 US gallons per hour equates to 0.7 US gallons per minute. 31.2 total capacity divided by 0.7 gallons consumed every minute gives 44.57 minutes. AH gives 43 minutes total which is pretty good (compare with the AH figures derived from AH's total time).

War Thunder however has the 21 minutes of total WEP too short. They give a consumption rate of 5.64 litres per minute, or 1.48993 US Gallons per minute which would be 89.39 US gallons per hour. So they've got their consumption rate wrong by just over twice the historical rate of consumption.

[icepac]

Yes, water/meth. injection is often used to eliminate inter/aftercoolers in race planes but they likely used both in warplanes.

The reno race planes actually end up using a much smaller ratio of alcohol vs water than used in the war because they ran into too much cooling of the intake charge.

It's the evaporation of the alcohol that causes such a huge cooling effect.

[nrshida]

Back to the Ki-84 for which I am arguing for an increase in WEP time from the increasingly suspicious 10 minutes total duration (in AH3).

We do not have reliable consumption rates for the Homare (the Ki-84 and N1K's engine). However we do now have a fuel consumption rate for the Bf109K-4, balls to the wall, full WEP:

141 Imperial Gallons per hour = 169 US liquid Gallons per hour on full boost and WEP.

For gasoline fuel, there is such a thing as stoichiometric air–fuel mixture. An ideal ratio of air to fuel that burns all fuel with no excess air. There is minor variations of lean / rich, ‘air-cooled’ engines tend to run a little richer to aid cooling but the typical stoichiometric air–fuel mixture is about 15:1. This isn't an engine-design consideration. It's essentially a chemical reaction one. A 'standard'.

The Homare and Daimler-Benz have comparable displacements. The DB is liquid-cooled, the Homare air-cooled, the DB a four valve head, the Homare two-valves per cylinder. The Homare does of course have 50% more cylinders which offsets the valve curtain area. Both are supercharged. The supercharging essentially means the 'fluid' is flowing faster through the hosepipe but the hosepipe has a fixed diameter.

It is extremely unlikely that the Homare's MW consumption rate could have been significantly higher than the DB605. It is possible in fact it was lower if the manifold pressure is lower, the mixture was richer or the specific output was lower. You just cannot push endless quantities of MW into the combustion chamber. There is an ideal, stoichiometric, ratio and I propose the data for the DB605 illustrates what that is:

DB605 - 35 gallons MW per hour and 141 gallons fuel per hour suggests a stoichiometric ratio of MW to petrol at a 1-4 (MW-fuel ratio).

Given the MW tank in the Ki-84 was larger, it is reasonable to assume that the Ki-84’s WEP would last at least as long as the 43 minutes total AH3 gives to the Bf109K-4 (which seems quite historically accurate).


Ironically, given the MA fuel consumption and capacities the unlimited system of AH2 was probably accidentally more accurate than what AH3 now has. If you are going to guestimate figures, please use reasonably similar aircraft's systems as a starting point. There are already game handicaps in place to balance competitive planes, if an incidental handicapping system is the intention.

[icepac]

Dago Red, a reno racer uses most of a 60 gallon tank of "adi" (water/alcohol) during a 15 minute race.

It puts out around 4000hp at 120 inches (possibly up to 150 inches) of manifold pressure and uses around 25 gallons of the mixture injected into the engine during a 15 minute race.

They also have spray bars that spray around 25 gallons onto the coolers during the race.

I just don't see an engine making half the horsepower of Dago Red using 5 times as much meth/water mixture for the same time period.........and they aren't spraying half of it on the radiators.

I agree with nrshida.....strange things are afoot at the circle K.