Jump to content

Holtzauge

Members
  • Content Count

    1034
  • Joined

  • Last visited

Community Reputation

746 Excellent

1 Follower

About Holtzauge

  • Rank
    Member

Profile Information

  • Gender
    Not Telling

Recent Profile Visitors

1220 profile views
  1. @Chill31: I think unreasonable's idea above about a separate thread for your test results sound good: IRL flight test data from a Fokker Dr1 deserves a thread of its own. Will also be so much easier for people to find rather than if the info is hidden away in a thread about damage modeling.
  2. @unreasonable True, IRL combat flying would probably not have included much level turning but more 3D manouvering. I just watched Mikael Carlson's Dr1 flying at Hahnweide and it look like you need to get the nose down when you manouver to keep the speed up. Here is an interesting interview with Mikael Carlson. Note what he says about looping: You are almost out of speed at the top! The level turn comparison I'm hoping to get from Chill is for comparison only to tweak the simulator FM. Once the model is tuned other more interesting flight conditions can be simulated.
  3. @Chill31: Here are some initial C++ simulation results for a Dr1 at 757 Kg, 79 hp assumed at an altitude of 3000 ft STD at 15 deg C (I don’t know how much the Le Rhone loses with altitude from 85 hp at SL but I’ve assumed 79 hp at 3000 ft STD): Max momentaneous turn at Clmax starting off at 85 mph IAS keeping altitude but letting speed drop: 13.5 s to do a 360 deg turn. Rough estimates based on the max initial turn rate at 85 mph IAS when entering turn for a 360 degree turn keeping speed up by dropping in altitude: circa 7.2 s. Stationary turn time at 55 mph IAS keeping altitude: 18.5 s to do a 360 degree turn. BTW: I found the text below on the Aerodrome forum by someone with the handle baldeagle which includes rolling in and out of the turn in a “Triplane” (I’m assuming a Dr1?) indicating I’m currently a bit pessimistic in my 18.5 estimate which does not include that: “By the way, be careful where you get your numbers from, the last Triplane I flew I timed a 360 degree turn, from level to level again, so including rolling in and rolling out, and it was almost 20 seconds, even with Voss and a rotary engine it isn't going to be much less, certainly not anywhere near 5 seconds. Those figures you get from 1917 are measured by very uncertain means, not to mention the exaggerations made by manufacturers, pilots, and anybody with a point to make. I wouldn't put too much stock in contemporary figures, except in very general terms.” Any thoughts on baldeagle’s comment above? Late edit: Just realized based on @ZachariasX and @unreasonable comments below that it would be good to measure the turn times going both left and right to capture the gyroscopic effects.
  4. I feel a need to clarify what I meant by the you in my original post: The "you" was of course me a no one else! Qualifications mean nothing if you go against the grain.
  5. Well I'm an engineer and I think you should go right ahead because nobody listens to to you anyway! On a more serious note: The nice thing about these forums is that there are a lot of knowledgeable people around: One of the best informed on WW2 aircraft performance I've come across is actually by profession a dairy farmer in Finland!
  6. @Chill31; That video you posted is just superb and your plane for sure is a sight for sore eyes. I guess many would have painted it red but the dazzle scheme you have chosen looks dandy! About the realism of sims compared to real life: IMHO the sometimes derided “on rails” FM is more realistic: Granted, my IRL experience is limited to gliders and light planes but I have also studied a number of good videos from the cockpit position both in the Me-109 and a Yak etc. and it’s very clear how solid and precise the response is to stick input with no tendency for the rubber band behaviour we sometimes see in sims. Il-2 has gotten very much better over the years and the change that was made to the Me-109 FM a while back was definitely a step in the right direction I think. In Il-2 the Fw-190 FM looks and feels the most realistic to me. The ground handling is a chapter in itself and I guess we just need to accept that this is difficult to model with today’s state of the art flight sim technology. I made some pen and paper ballpark calculations about the effect of rotary engine gyroscopic effects compared to elevator/rudder authority a while back and I have a hard time accepting that it is so difficult to counter that IRL as it is in many sims so would be interested to hear your opinion on that. Regarding the turn time comparison: I have no way to compare with your measured times unless the altitude is kept constant during the turn: Any digging will skew the results and there will be no way for me to take this into account. This applies for both the instantaneous and constant turn so if you could keep the altitude during the turns relatively constant then we can compare, otherwise it will be difficult to draw any conclusions I think. About the 24 by 6 foot wing: For sure I can simulate that but why? Where did you get those figures? That would result in an aspect ratio of 4 and that is higher than what I’m assuming. Also, that is a lower wing area than I’m calculating with. I can hopefully get some simulations done tomorrow but it won’t be tonight because me and the missus have just polished off an excellent bottle of Cotes Du Rhone and the rest of the evening is off!
  7. OK, good, then I have the data I need. I could run a simulation before you fly based on the info you gave. In that case, what altitude do you want me to run it at? I’m assuming you don’t want to be too low if you are going to push the limits? I can run the simulation at any altitude and turn entry speed (IAS) so if you give me those numbers then I can set it up. If you have imperial gauges in your plane just give me the knots or mph IAS and altitude in feet and I will set that up in the simulator. So far it looks like I’m much more conservative than FC: For example, in FC the best I manage is around 7 s for an instantaneous 360 degree left turn from 160 km/h IAS at 7-800m. At the same altitude I can do continuous turns (left) at 9 s with about 100 Km/h IAS. In my C++ simulation however, I get 9 s and 12 s respectively for the same manouvers assuming a 122 hp Oberursel and when I run the simulation with 85 hp the numbers will of course be even lower than that. In addition, I have nowhere near the energy retention in C++ as in FC: Doing max rate turns is basically like hitting a brick wall in my simulation: The speed drops dramatically fast and I can not hold g’s as long as in a turn in FC. I have tuned the Dr1 C++ model as best as I can given the data I have but there are two things which involve a lot of guesswork and that is the equivalent aspect ratio to use and the Oswald factor e. Alas, I have no data for equivalent aspect ratio for a triplane but for a biplane Hoerner’s book Fluid Dynamic Drag provides some guidance in figure 19 page 7-12 which I have used as a base. BTW: before anyone in the forum blows a gasket for me “trying to nerf” the Dr1 rest assured: I have modeled the Camel in C++ and it’s not only the Dr1 that is different in FC and my simulation. They are both off. Question I’m trying to determine is am I too conservative or is FC too optimistic that’s all. PS: Forgot to add: The simulator uses standard atmosphere model 15 deg C and 1013 mbar so when determining altitude if you could set your altimeter to STD that would be great.
  8. Well 20 rounds fired at the same time by a steady hand with two planes in low relative velocity to each other hitting the spar in a fist size pattern ought to do the trick for a 1 g wing loss if that is what you are trying to capture. I was more thinking of a few passes adding up to 20 hits in the box meaning the rounds are more evenly distributed as more representative in the majority of the cases. Seeing the spars are typically dimensioned for more than 5 to 6 g's you would need quite a concentration of hits to weaken the spar to the point of a 1 g break occurring and I have a hard time seeing that occurring in 45% of the cases in a 1000 sample which would never occur in the DM model I'm suggesting. Anyway I think we have to end by agreeing to disagree on this one.
  9. Regarding the SE5: Just because a stress analysis shows a problem and suggests a reasonable solution does not mean it was actually incorporated and the same standard of evidence should be used here as applied for the Fw-190 Clmax (The BOS Wurgergate affair): Keep it at the measured value until proof to the contrary is found. No double standards! About the figures you linked to above I see a major problem: The risk of losing a wing at 1g and 5 g are far too close: I think this way of counting could be improved and I see I have failed to properly explain what I meant earlier on: My suggestion is somewhat different: I suggest you adjust the STRENGTH of the airplane depending on the number of hits, not the probability of losing a wing at a certain g-load meaning if you stay below the wings reduced strength you have 0% chance of losing the wing below that g-load. First, my idea for the insta-failure/golden BB fail works like this: As I understand it the wings in-game are divided into circa 10 hit box sections each. Now each of these sections would have two possible scoring systems depending on what type of structural members they include. For example, most of the hit box sections only contains spars, ribs and fabric. However, some sections also incorporate struts and bracing wire. Now the sections that contain struts joints, bracing wire, i.e. single point of failure parts would toss the dice for a complete loss of the wing irrespective if the g-load is 1,2,3,4 or 5 g. If the dice says your number is up then it’s up. However, seeing the target area is so small I suggest something like a 100 hits in that box area for a 50% chance of a loss, i.e. 0.994 chance in your favour of avoiding a KABOOM fail every time that box is hit. Next, the accumulated DM: Boxes without these type of structural members only log the second part which is accumulated damage as listed below. The hit box begins at full undamaged g-load rating say 6 g as an example. Every time a bullet hits (parameter n is stepped by 1) that sections strength goes down like this: Tolerable g-load= 6*(0.965**n) So after 20 hits you have cut your strength in that section in half, meaning you can pull 3 g and get away with it 100% of the time. If you get 5 hits, you can hold anything below 5 g etc. Personally, I would add 5 "free" hits or something similar that are not counted before I start stepping n because dropping directly down to 5.79 g with the first hit seems excessive. Anyway, all this can be tuned of course and it would address the current DM’s 45% chance to break a Camel wing at 1g after only 20 hits which I think is not that good a solution. Nota bene: The numbers I have suggested above are just that: examples and the DM model above is highly tunable, both in terms of how likely things are to occur and is of course potentially differentiable between different aircraft if one would like to do that. In addition, if the current DM keeps track of the hit area depending on the plane of attack all the better since this could then be used to tweak the numbers I suggested above.
  10. I have no idea if they did beef up the structure after that report came out. Anyway, it looks like the undamaged in-game SE5 handles a lot more than 5.5 g. At least that's my impression from pulling the wings off a few times now with the g-meter we have. However, everyone seems to be focused on battle damage right now though and for some reason are happy with in-game undamaged g-load modeling even though that seems to be optimistic. At least that is my conclusion based on the two NACA reports I posted but I'm certainly willing to be proven wrong if someone can come up with the evidence to support that. To be a bit provocative, looks like some here are on a quest to find evidence that supports their style of flying rather than what is historically accurate if you get my drift..... That being said I have no idea if some of the other planes like the SPAD for example are more weak than others or the D7 is stronger than it should after taking damage. OTOH assessing the actual differences between these in terms of what is reasonable can only be done with very detailed modeling of the spars and logging exactly where the hit occurs. There are advantages and drawbacks with single box spars, double solid spars, I-beam spars etc. and how do you model that fairly without a very detailed model? Since that is never going to happen I think the best first order approximation is to differ the designs according to how strong they were undamaged (e.g. breaking g-load 5, 6, 7, 8 g etc.) and then deduct the same amount battle damage points from all planes when hit box is struck. The DM count should be split in two parts: one that every time rolls the dice for a catastrophic failure per hit and one that accumulates damage. Again, IMHO the latter should step very slowly since most part of the hit boxes in the wing are just ribs and fabric meaning we should see very few planes shed wings in-game due to battle damage and not see any marked difference between say the SPAD or D7 in terms of how sensitive they are to being hit in the wings.
  11. Yep, for sure: In WW1 there was probably a lot of variation between individual aircraft seeing that it was more furniture building if anything and that some hooves made better glue than others. In fact our indigenous Swedish “mini Fw-190” the J22 was in fact built to some extent by furniture companies because SAAB had other fish to fry at the time. Back on subject: According to NACA report 143, ANALYSIS OF STRESSES IN GERMAN AIRPLANES By WILHELM HOFF (looks like a very interesting doc BTW) "Bau- und Liefervorschriften der Inspektion der Fliegertruppen" (BLV) determined load regulations and was issued in 1915, 1916, and 1918. On page 19-20 the subject is wings and as far as I can see skimming through it is that the pullout load requirement was initially 6, the 4.5 for a time to end up at 5 g in 1918. This was the limit load factor as I understand it meaning the design passed if it could handle this and nothing more. Interestingly, these German design rules tab well then with the SE5’s wings breaking at 5.5 g meaning that our in-game crates are actually stronger than these numbers. I can see you all enthusiastically waving your arms suggesting I send this report to the developers asking them to lower the wing strength of the scouts in game so we can see more wings snapping of both in SP and MP. Seriously, what this tells us if anything is that if our in-game scouts can handle 7 and some even 10, we should be very happy. In fact this is probably quite realistic as well: I have done structural design myself and the last thing you want is for your 5 g rated design to go KABOOM at 4.99 g in the end acceptance testing and structural designing every piece to break at 5.00 g is impossible meaning you always include a safety factor on the safety factor so if the limit load is 5g then getting to 7g before the wing actually fails when sandbagging it in the static load acceptance test would probably not have been uncommon. All in good time! But before that it would good to know @Chill31 Dr1 data: What is the weight and what is the power output? Also, is the wing profile the original Göttingen or some more modern design? In addition: About the "not bleeding energy part" I managed 9 s sustained turns in the Dr1 in game while in the C++ simulation I get around 12 s to do a 360 degree left hand turn at around 800 m altitude so it would be really nice @Chill31 if you could give us your Dr1's best sustained turn time for a specific altitude of your own choosing so we could compare?
  12. Too bad! Was looking forward to seeing those results. I have developed a C++ simulator and actually modeled the Dr1 so seeing some IRL data would be nice. According to my C++ simulation, if you start off at say 160 Km/h TAS in a Dr1 at 750 m altitude and do an instantaneous max rate turn it takes about 9 s to do a 360 degree turn. Then, after the first 360, the speed has dropped to the speed for best turn rate which I estimated at about 80 Km/h at which time the turn rate is steady at around 29 deg/s or about 12.4 s to do a 360 turn. Interestingly, in my C++ simulation the g-load is 3.8 g at Clmax about 1 s into the simulation and then rapidly falls to 1.5 g riding the edge of stall so nowhere near 6 g's. The weight I used in the simulation is 575 Kg which maybe is a bit different from your T/O weight? In addition I'm assuming that the Oberursel delivers 122 hp SL and I don't know what power output you have in your Dr1? Anyway, super interested in your test results since I hope to be able to use them for tweaking the C++ model! About the Dr1 6 g limit: Is that the max allowable limit meaning the design is stressed for 1.5 x 6 =9 g or is the max allowable 6/1.5=4 g?
  13. I actually work with 5G nowadays so I'm not planning to visit Glastonbury in the near future. I can tell you it was a challenge to get that particular feature into the system since it initially affected download speeds but with a bit of tweaking we got it right in the end.
  14. I know, for some reason I keep blacking out at 89.9 degrees but maybe I'm not straining enough.
  15. I always find it hard to hold the turn coordinated at 82.26 bank angle and I sometimes end up in a 82.78 degrees or 11 g turn but then I need more practice I suppose......
×
×
  • Create New...