Aircraft ROC Review

[Dili]

The Elf since you seem to know the guts of Air Combat can you tell what data should go to build the elusive MVR(Maneuverality)? Since CLIMB doesnt count for air combat then it should go to build MVR (or power to weight ratio) and what about speed?

[el cid again]

IF ROC is "not used by bombers" ( and presumably that is shorthand for "all non-fighters" ) - then it cannot hurt to get it right. Further -

player use the data as a reference - and nothing wrong with it being right for that purpose

further

IF the data is right, and code ever DOES us ROC for other purposes (or indeed, if it uses it right now in ways you don't know about) - well it is still right.

Finally - if we are using ROC as a component of maneuverability - we still need to know what ROC is to some standard.

I am sorry if I have confused you about what a "standard altitude" is. It is an altitude the time to which we would use to determine aveage ROC. The same plane has a DIFFERENT average ROC to 5,000 feet, 10,000 feet, 25,000 feet, 3,000 meters, 5,000 meters, 10,000 meters, name an altitude - the average ROC is different. Which one is "right" by your lights? How can it ever be right to use A6M2 average ROC to 5000 meters, but F4F average ROC to 10000 feet? We need to have a common standard.

I think using initial ROC is a great standard. Aside from the fact it is sometimes available, it is also a true performance point - like maximum speed or range or ceiling are. It is a way to make aircraft different. All we do here is push electrons - we need to make each package different - or they are just labels for generic planes. IF ROC is indeed used to determine who has the bounce, should not the best performing plane be the one biased by the routine and data to get it? I think we are on the right track after all.

[akdreemer]

ORIGINAL: Dili

The Elf since you seem to know the guts of Air Combat can you tell what data should go to build the elusive MVR(Maneuverality)? Since CLIMB doesnt count for air combat then it should go to build MVR (or power to weight ratio) and what about speed?

Maybe also altitude, roll rate, along with speed. At higher speeds the Zero had real problems maneuvering due to increased airelon stiffness and speed increased.

[JeffroK]

The old "Europe Ablaze" game had a modifiable database.
 
Amongst the aircraft rating were "combat ratings" for Low-Medium-High levels. This allowed you to show the differences in performance for each type, a P39 might rate 4-3-1 , a Spit VB 6-4-2, bf109G 6-7-9 . The rating took into effect all of the variables like ROC, Mvr, Dur (IIR Top speed & Cruise were seperate)
 
You sent your raids in at the 3 levels and  the interceptors would have to change heights, sometimes the 2 level difference would enable you to sneak in.
 
To try to replicate every aircrafts performance at every height level would be pretty impossible, dont forget weather would have an effect etc etc. Whichever figure is used, it would have to be consistant for each model, and not throw out the a-a combat engine which has been done by mods to firepower or durability.

[el cid again]

ORIGINAL: TheElf

ORIGINAL: el cid again

DILI: OK - I see what you are trying to say. And you wish to use ROC to define that??? Interesting idea. I am not sure we can do that. Stock defined altitude in hard code for two cases. RHS took that out (by defeating the slots - putting in planes that won't fight up high anyway) and replaced it with an operational altitude system - we don't let the bad planes go as high. We also put in power loading and wing loading as factors in maneuverability, which should help some planes be better than others. IF this is not going far enough - we could further modify maneuverability - which is where for sure ROC is being used in every air combat (at least in mods that make it a componant of maneuverability).

Can you identify ALL such aircraft? Can you identify them with some QUANTIFIABLE rating? IF so, I am inclined to work out a special case for them (as we do, say, for the geodesic construction of a Stirling, or the cast armor nose of a Sturmovik re durability).

The problem with using average ROC to identify which planes are poor and which are good at fighting is that it penalizes the good planes - not the bad ones. Sort of the opposite of your goal I would think. Unless you believe "a good ROC is always an indicator of a poor plane in a fight." Which seems unlikely.

code does not use ROC in Dogfight. Sorry.

Well - yes and no. It probably does not directly - in the sense of using the ROC field as an input for the routine. But it does use maneuverability. And to the extent ROC is a component of maneuverability, it is. It does not really matter if that was the original intent or not: if it was not the routine has been "tricked" into using it indirectly. And because that is so - minimizing the value of ROC means you are minimizing the impact it has on each and every combat round.

But these comments - yours and mine - are ignoring a different truth: we are told there IS a case where ROC IS used in air combat - determining who gets the bounce. I believe it is likely there is yet another case - the case that fighters not on CAP must climb to intercept an incoming raid. [If not, there should be, and probably someday there will be] Anyway - here - for determining the bounce - the ROC IS used - directly - as such. And again, I fail to see why we don't want to recognize the better performing fighters at doing this?

[el cid again]

ORIGINAL: TheElf

ORIGINAL: TheElf

Regardless - we want data to a consistent and defined standard. Average ROC is not nearly as clear a standard as it sounds like: Do I use the average ROC to 25,000 feet (the one actually listed) for a B-29? Is that fair to the B-29 when it is at 5,000 feet? Even if you limited the idea to a fighter - why should a high altitude fighter get a seemingly low ROC when it is three times better than that at lower altitude? Unless we adopt a standard for average ROC - we still have a mess. IF we adopted a standard, we have to convert a majority to it, and we have to decide what average makes sense?

Bombers do not use RoC.
Do what you like, but the code does not care.

In the first instance, my assignment is "get the data right." At a minimum it means players looking at comparative statistics will have meaningful numbers to look at.

In the second place, we don't know where a field is used? Just because routine A - or routines A and B - do not use a field do not mean that routines M, X and Y don't use it.

In the third place, the value used for ROC is used indirectly by every routine that uses the maneuverability field.

In the fourth place, we cannot know what code may do in future? But being ahead of the code - if we are - is never a bad thing. It has happened before.

Finally, we may some day want to do the routines ourselves. There have from time to time been suggestions that WITP I might become public domain in some circumstances. Alternatively, the data bases might be used with a different engine - say Vassel (another Matrix product). Having the data needed to model well is not going to hurt us in any contingency.

[el cid again]

ORIGINAL: AlaskanWarrior

ORIGINAL: Dili

The Elf since you seem to know the guts of Air Combat can you tell what data should go to build the elusive MVR(Maneuverality)? Since CLIMB doesnt count for air combat then it should go to build MVR (or power to weight ratio) and what about speed?

Maybe also altitude, roll rate, along with speed. At higher speeds the Zero had real problems maneuvering due to increased airelon stiffness and speed increased.

How do you put altitude into maneuverability?

How do you get roll rate data? We tried to talk about this in a thread - with a very senior mathmetician - but there is no really easy answer about how to calculate it. And it is not generally available in most references. It matters not how nice it would be if we could get it accurately if we cannot get it at all. How do we get it? And IF we get it - how do we use it?

Speed seems to be already in every variation of the maneuverability field. Sometimes it is the ONLY component. Other times it is the majority component.

What I wonder is if there might not be a way to get rid of speed and what you called "altitude" and put in energy? This might be the sum of momentum (speed times mass) plus the potential energy of position (that is, altitude, with mass and the acceleration due to gravety as factors).

For clarity sake, let me note that RHS introduced adding (actually subtracting) wing loading and power loading to the product of speed and ROC. Further, RHS has increased the relative value of ROC by 250% - so instead of being the vast majority of maneuverability, speed is just a slight majority of the factor. Higher wing loadings and/or power loadings decrease the natural product of the two primary determinants, which are speed and ROC. ROC matters because of its impact on mitigating loss of altitude due to turning during combat. It appears that the base system used maneuverability based either mainly on speed times ROC, or on speed alone (some stock data fit both patterns). [This is the basic, single engine case. For multiple engines, RHS divides by the number of engines. (We would treat coupled engines as one; we would treat fore and aft centerline engines as one; but we have no instance of either). Stock and CHS divide by 2 for twin engines but by 8 for four engines - for reasons unclear to me.]

[el cid again]

I am finding a large fraction of Allied aircraft already use initial ROC data. It seems to be a majority of US Naval aircraft (EDIT to this add RN aircraft). The US Navy (and the FAA) apparently consider initial ROC a major technical perameter - and it is virtually always listed - while "time to altitude" data is almost never listed. So - in a data set where people took any sort of ROC data from whatever source was at hand - we got an unintended bias: USN virtually always got credited with initial ROC - IJN never did. There is also a third case: where average ROC data was used. Here the bag was also mixed. Typically Japanese aircraft are listed with "time to 5000 meters" (16,405 feet) while Allied aircraft are listed with "time to 10,000 feet." The average ROC of an aircraft to a higher altitude will always be less than to a lower altitude. So the same prejudice (statistically speakint) is in the data where average ROC was used. Arguing that no standard is required, and we should leave well enough alone,
is to rationalize this built in statistical prejudice - conscious or not. It may or may not be better to use initial ROC data: but it appears that possibly a majority of planes already have it - and certainly many do. We would have to do a lot of work to change that. And many of the average ROC cases will not be to a standard altitude - if we adopted one - so they too would mainly have to be reworked.

And I am not hearing any suggestions about a standard altitude. I am not negotiable about having a defined standard.
EITHER we use initial ROC, OR we use average ROC to a defined altitude. Just collecting any sort of ROC data and pretending it is relatively valid is not reasonable (although it is economical in time - and if that time is paid - in money).

[m10bob]

I don't know if maneuverability is a combination of influences, to include ROC.
If so, it makes sense.
I do however see the need to change the ROC on most planes in game, and have provided technical source reference for this info.
I have never flown a stove, but I understand they can fly nearly straight up a long distance and never lose speed.
Marvelous.
Fans(props) can't.
In the era we are simulating ROC is a more important factor than to relegate it to insignificance with small representation (or inaccuracy) in-game.
The air space of an air battle in-game is spread over a 60 mile hex, but the ability of a plane to turn is still represented *somehow*.
Why then would we not seek written and verifiable ROC info and correct "service ceiling"?
Maybe the program is smart enough to "read/interpret" the info we put in those boxes (in the editor), as much as it can read the dates of production??

Since a lot of WITP mods have been fairly experimental, I suggest we just put that verifiable info in the provided boxes and see what happens?
I have doe this with my aerial mod and I personally see a difference.
Of course I am prejuidice.
Let others see it and try it..
As for "turning ability", I provided the formulae for figuring power to weight ratios and wing loading early in this topic.
It works, and has worked for wargamers since at least the early seventies.
The info calculated is not measured from models, but from the real aircraft.
That info is then scaled to the model..

Here is the thread giving an outline of how to figure out the maneuverability of a plane..

http://www.rdrop.com/users/hoofj/methods.htm

Anybody with an old copy of "MUSTANGS AND MESSERSCHMITTS" has IMHO the most detailed example that even a dolt might comprehend.[;)]

[el cid again]

Me-109T initial ROC was 3346. This increases maneuverability to 33 (up from 30). This data from Green, Gunston giving the E model (T was a variation of E, and Japan had a licence for Es), but the T was even faster climbing.

[el cid again]

ORIGINAL: m10bob

I don't know if maneuverability is a combination of influences, to include ROC.
If so, it makes sense.

One Matrix programmer wrote that the "maneuverability" field may be "somewhat misleadingly named." Another wrote that it is the primary input in the actual air air combat routine. I suggested above we might wish to think of it as the "air combat value" field - not in terms of armament or protection (for which we have devices and "durability") - but in terms of the ability of an aircraft to engage/evade engagement. If this information/evaluation is correct, it does not matter what the design intent was, or what the maneuverability field was based on in stock. We can make it include ANYTHING we wish that affects the ability of aircraft to maneuver in combat - and the routine will use it - as a way to distinguish between aircraft - just as it does durability and weapons devices.

The one thing I noticed above which has my sincere interest was the suggestion that we include turning rates. I have never been able to find a good indication of this in basic aircraft data: the sort of information we have on most of our 250 types of aircraft. I think it is somewhat related to wing loading and also crudely modeled by ROC (in the sense that an aircraft with a horrible ROC is likely to turn like a dog as well). But I would not mind modeling it more precisely - if a way exists to easily and quickly derive it from basic perameters (e.g. wing area, weight, speed, etc).

[el cid again]

Corsair III/IV - using FAA data - needs complete revision. It is a classic case where the British used an "identical" aircraft in a very different way. It is heavier, armed for offensive work (with 2x 1000 pound bombs), and has a slower ROC as a result. It cruises at a higher speed for less time, which probably changes the range (261 mph for 240 minutes).

Max speed 415 mph, initial ROC 2070 fpm, maneuverability = 29.

I am treating this as eratta and working it into all levels for x.782 microupdate. What is the point of having separate aircraft types if they are rated the same? [Answer: there are still different pools and production rates, but I PREFER if there is also a difference in armament/performance]

Swordfish I initial ROC = 560 fpm does not change maneuverability = 7. This unusual value for a single engine aircraft is in the range more common for twin engine aircraft. A Swordfish on an attack run, seen from dead ahead, appears to "hang in the sky" - and its horizontal movements (wobbling side to side) are more noticable than its forward progress. By 1941 it is wholly obsolete for such a task, but it is fitted with early radar, rockets, and is regarded as an ASW aircraft and a "secret weapon" by Adm Coyningham - because it can find enemy ships/carriers at night - and lead attacks on them. Durability should be 12 (up from 10), endurance should be 346 minutes (down from 401) and both are being treated as eratta (ie changed in Levels 5 and 6). FAA data.

Albacore for some reason was unmodified since CHS. It was completely wrong in every possible sense except maximum speed. It should read Max speed 159, cruising speed 113, endurance 538 minutes (494 actually, we use a higher value to get correct operational ranges), initial ROC 780 fpm, maneuverability 9 (down from 16), durability 12 (down from 27). This will be treated as eratta and backfitted. FAA data.

Firefly is yet another case that is changing dramatically. It should be classified as a fighter, not a fighter-bomber. It has an initial ROC of 1800 fpm (FAA data, Gunston gives 1700 for the I model we use - and I am giving the Allies the benefit of the better rating). That still reduces maneuverability to 22 (down from 24). It also carries 2 x thousand pound bombs, just like the FAA Corsair did.

[m10bob]

"The one thing I noticed above which has my sincere interest was the suggestion that we include turning rates. I have never been able to find a good indication of this in basic aircraft data: the sort of information we have on most of our 250 types of aircraft. I think it is somewhat related to wing loading and also crudely modeled by ROC (in the sense that an aircraft with a horrible ROC is likely to turn like a dog as well). But I would not mind modeling it more precisely - if a way exists to easily and quickly derive it from basic perameters (e.g. wing area, weight, speed, etc). "


Both turn ability and acceleration can be calculated by knowing HP vs weight and you can obtain the wing size in a simple google search, as I mentioned in a prior thread.
Just type in (example) "Brewster Buffalo wing sq ft", and you will get an answer..
(Short cut, sometimes just punch in "wing area" for the plane..)
Divide the wing by the hp.
Knowing the top speed and the stall speed gives you a range..stall 70, top speed 370..Range of speed variation is 300 miles..If you know you lose "x" speed in 5 seconds from a turn, you can figure how many 5 second increments you can turn before you stall.
The plane which turns longest turns better..
Therefore, an A6m2 weighing 1/3 that of a P 47 and having half the HP can turn 3 times quicker..(approx)..
Of course a P 47 would be nuts to try to turn with a Zeke..

[el cid again]

Well - we do have wing area - hp - and weight - but we do not have stall speeds - which don't seem too critical. Further - we have weight and wing area in wing loading, and we have hp and weight in power loading - so maybe we have these already as indictors. But we are not dividing wing area by hp. Should we?

[m10bob]

ORIGINAL: el cid again

Well - we do have wing area - hp - and weight - but we do not have stall speeds - which don't seem too critical. Further - we have weight and wing area in wing loading, and we have hp and weight in power loading - so maybe we have these already as indictors. But we are not dividing wing area by hp. Should we?

Helps determine the stall speed and turn abilty..
See the formula in that link in my post above..

[Dili]

Albacore was worse than Swordfish in almost everything. So much that Swordfish continued to be used and Albacore its supposed replacement was dropped.
 

It is heavier, armed for offensive work (with 2x 1000 pound bombs), and has a slower ROC as a result. It cruises at a higher speed for less time, which probably changes the range

 
Can you explain how a fighter bomber with 2x1000lb bombs have a better cruise speed?
 
 

[TheElf]

ORIGINAL: Dili

The Elf since you seem to know the guts of Air Combat can you tell what data should go to build the elusive MVR(Maneuverality)? Since CLIMB doesnt count for air combat then it should go to build MVR (or power to weight ratio) and what about speed?

No.
Speed is a separate variable.

Try not to combine too many variables into one value. It tends to be too one-dimensional.

[el cid again]

Dili: Sure. Cruise speed is a choice, not something ordained by physics (as, say, maximum speed is). What you gain or lose is efficiency. If you choose a higher cruise speed, usually, you will lose range, because fuel efficiency declines. This is why the same aircraft in FAA service does not have the same range as it does in USN (or RAAF, or RNZAF) service. IRL there are other aspects of this matter: altitude for example. In the USN material on aircraft, you find that speed is specified BY ALTITUDE, and you get different maximum or cruise speeds if you make different altitude choices. The choice one makes for a naval aircraft may not be what is optimal in engineering terms: in that era (pre electronics) visual search was a key requirement - and so altitude for a naval aircraft might be chosen for operational reasons. In the present age, with primarily jet aircraft, altitude makes a whopping 300% difference in terms of fuel efficiency - and a significant difference interms of maximum speed (jets love to fly higher). In WWII, only some planes could get high enough to experience the lower end of these effects (again, higher altitudes were more efficient, but not by a power of three). OTH, a truly low flying aircraft may gain efficiency from ground effect (which term applies over water even more than over ground, because you often dare fly lower). Anyway - in a simulation - we should be using the INSTITUTIONAL data - the choices of FAA ought to apply to their aircraft - and USN to theirs.

And your history (re Swordfish and Albacore) is dead on. Every source says something similar. The Swordfish was one of the most amazing of great aircraft - enough so I wonder if we should not mess with its durability a bit?

[TheElf]

ORIGINAL: el cid again

IF ROC is "not used by bombers" ( and presumably that is shorthand for "all non-fighters" ) - then it cannot hurt to get it right. Further -

True...it's your time, but think about this...how do bombers usually fly in combat?

player use the data as a reference - and nothing wrong with it being right for that purpose

further

Books are used for reference too.

IF the data is right, and code ever DOES us ROC for other purposes (or indeed, if it uses it right now in ways you don't know about) - well it is still right.

IF the data is right, and the code ever DOES use proper tire pressure...

Finally - if we are using ROC as a component of maneuverability - we still need to know what ROC is to some standard.

You can use it as a component of MVR if you want, it's your mod, but it isn't used in the dogfight code because it is used elsewhere. So why would you want to combine it into MVR if it is already being accounted for? Using Power to weight is the same thing as using ROC, it's just expressed in a different format. I suggest sticking with that. You are over thinking this.

I am sorry if I have confused you about what a "standard altitude" is. It is an altitude the time to which we would use to determine aveage ROC. The same plane has a DIFFERENT average ROC to 5,000 feet, 10,000 feet, 25,000 feet, 3,000 meters, 5,000 meters, 10,000 meters, name an altitude - the average ROC is different. Which one is "right" by your lights? How can it ever be right to use A6M2 average ROC to 5000 meters, but F4F average ROC to 10000 feet? We need to have a common standard.

I know. But what is a standard altitude in WitP?

I think using initial ROC is a great standard. Aside from the fact it is sometimes available, it is also a true performance point - like maximum speed or range or ceiling are. It is a way to make aircraft different. All we do here is push electrons - we need to make each package different - or they are just labels for generic planes. IF ROC is indeed used to determine who has the bounce, should not the best performing plane be the one biased by the routine and data to get it? I think we are on the right track after all.

But as Dili said comparing a/c by their initial rate of climb yields a skewed relationship compared to using an average ROC to SC. You are obviously aware of this since you said as much above.


[/quote]

[el cid again]

ORIGINAL: TheElf

ORIGINAL: Dili

The Elf since you seem to know the guts of Air Combat can you tell what data should go to build the elusive MVR(Maneuverality)? Since CLIMB doesnt count for air combat then it should go to build MVR (or power to weight ratio) and what about speed?

No.
Speed is a separate variable.

Try not to combine too many variables into one value. It tends to be too one-dimensional.

In spite of the fact NO ONE is doing this in WITP, I must say I agree with The Elf. About the only excuse I have for not going over completely to this point of view is that we cannot redesign the routines. At least we know that speed (and also cruising speed) are used at different points - even in combat related routines. But the maneuverability field DOES contain speed data. It may be it ONLY contains speed data - and that was the first suggestion I got from a programmer. SOME stock aircraft seem to use max speed/10 as maneuverability. I believe the original system was to combine speed/10 with ROC/500. By changing that to speed / 20 and ROC / 200, we changed the ratio - so I have moved in the direction of reducing the speed factor - if not getting rid of it altogether. Adding loading (wing and power) also diluted speed slightly more.

In an ideal world, we would go the way of The Elf: if we wrote the code we would separate speed and maneuverability entirely - and we would split horizontal and vertical maneuverability - possibly calling them ROC and turn rate. If we had the fields we might do this for (say) three different altitude levels (5000, 15000 and 25000 feet????). But we don't get to do that. [Where 5000 means 1 meter to 2000 meters, 15000 means 2001 to 6000 meters, and 25000 means 6001 meters and above]