The day had finally come for the Allure Bipe maiden flight. The usual pre-flight range checks were done.
I also double checked the direction of all control surfaces! It’s very easy to get something wrong there. It’s good practice to check this at the start of every flying session for your F3A model. Many have taken off with ailerons disconnected..
With all the pre-flight checks done it was time. Just like most maiden flights, the model just needed a click or two of Aileron and elevator trim to be flying perfectly. :-) If only that was the whole story... It’s true that minimal trim adjustment was required but a severe driveline vibration brought things back to earth quickly. Upon checking the upline trim of the model the vibration reared its head anywhere above 50% throttle. The model was quickly landed to check for anything obvious. Nothing obvious was found so to the air again. The result was no different in fact the vibration seemed worse. Upon further checking it was noticed that one of the rear support ears had separated from the fuse side. The second one had started to delaminated the glue joint. Bugger! That was the end of flying for that day. The saga begins.
At first I thought my normal rear support ears were inadequate for the prototype Adverrun Single. A set of 3mm AC ply ears were made, fitted and just for extra security had carbon cloth applied both sides of the ear.
Off to the field again. Unfortunately, the result was the same with severe vibration heard. The model was landed and I headed back home with my tail between my legs.. :-(
In hindsight, I should have first tried the drive at home before heading to the field. Ah well, we live and learn all the time.
More to come!
Prototype products can (and do) have their issues. I went into this build knowing that I could have some issues and as it turns out there were issues. On the positive side, it was a good learning experience. Below is a picture gallery showing some of the things tried and tested to eliminate the severe vibration I was experiencing.
I also checked the prop balance. It was perfect. However, when the prop was placed on a flat surface and the tip heights checked, I measured 2mm of variation. Not good. A second prop was checked and found to be ok. Time to test again. Unfortunately the vibration still existed. In fact the vibration was so severe that the prop lock nut came off the drive and hit the prop. There goes that propeller and nut. The leading edge of the prop was badly damaged and the nut had been flicked off to outer space. Not happy Jan! :-(
The second thing I looked at was what has changed on the production Single Drive. The main change was shortening the main prop shaft from 100mm to 86mm. Adverrun advised this was done to reduce vibration. Mmm, sounds familiar... Although I could shorten the main shaft quite easily, it would have required the firewall to be moved further forward which is quite major work. I decided to do some more checks before going down this path.
In a couple of the picture below you’ll see some white Delrin spacers. The rubber isolators on the rear support were removed and the rear of the drive hard mounted. On testing, it could be seen that this change had little to no effect on the vibration above 50% throttle. The boys at Adverrun suggested that all isolators could be removed to hard mount the whole drive. I was reluctant to completely hard mount the drive so I decided to check some other things first. They also suggested shortening the shaft to the same length as production drives.
The motor and belt were removed from the drive. With the shaft and pulley now able to freely rotate, I could visually See runout in the shaft and pulley. A dial indicator was used to check the shaft runout. I measured 0.07mm of runout. The shaft was bent. The pulley was removed from the shaft by removing the press-fit roll pin. The shaft was then placed in the lathe using a four jaw chuck. I was able to repeat the runout measurement. The pulley was also checked and I found the shaft hole to not be concentric. I had now identified a few possible contributors to my vibration. The prop, shaft and pulley. I then decided to make a new shaft. I happened to have some suitable 10mm steel for the job. I won’t detail the process of making the new shaft but can say that runout on the new shaft was less than 0.02mm. Much more acceptable. The pulley was mounted into the lathe four jaw chuck and bored out so a sleeve could be fitted. I now had a shaft and pulley that ran significantly truer. :-)
The Drive was reassembled and tested. The vibration still existed but now occurred at approximately 60-70% throttle.
Next! I was of the opinion that the distance between the front rubbers and rear support was too narrow. My theory was that moving the rear mount further aft would provide more mechanical leverage to the mount system. A new carbon plate was made to move the rear support back 30mm. During prior tests it was noted that the rear support ears would wobble or flex during vibration. A carbon brace was made to tie both rear support ears together. With this change the old ply ears had to be removed and repositioned 30mm further back. Although this change increased the vibration threshold to around 80-85% throttle, it still wasn’t a full fix. We were going in the right direction though.
Apart from the main shaft length, one other difference between my prototype drive and the production drive was the shore hardness of the front rubbers. I don’t think my drive came with rubbers fitted so Ralph Schweizer 10 x 8mm isolators were fitted. I only noticed this difference when handling both drives during a procrastination session. The production Adverrun Single uses 75 shore isolators whereas the Ralph Schweizer variety was around 50 shore. A significant difference. I decided to try the harder isolators on my prototype drive. There was nothing to lose.
Well, wouldn’t you know it full throttle without vibration could now be achieved! What a relief! The saga ends so now we can move onto flying again.
Now the trimming process continues...
The initial settings and throws for the Allure Bipe were copied from my Alchemy which included the downline mix. The CG was set as per Bryan’s recommendation to begin with. I follow Bryan Hebert’s trimming triangulation methods. If you don’t have a copy of his trimming guide you should consider the investment. The trim guide now includes Bipe trimming and setup.
With the model set for hands off straight and level flight, the up and down lines and knife edge performance was checked. The model went straight up with no pull to the belly or canopy. That’s a great start. In the downlines the model came straight down with no pull to the canopy or belly. Remember, I have a low throttle to down elevator mix. I was surprised that it didn’t need adjustment from the Alchemy settings. So far so good.
The knife edge performance of this model is pretty impressive to say the least. It’s certainly considerably more locked in knife edge than a mono. Bear in mind this is my first F3A Bipe. Back to the knife edge trim... Knife edge is checked in both directions. I found that in both directions I had a very small amount of pull to the belly. There was also a very small amount of roll coupling. At this stage we are just taking notes of what the model is doing.
The control surface trims were checked upon landing. All looked good with exception to the rudder. There was approximately 1mm of left rudder trim. This is compared against the fin tip. The same is done for the ailerons and elevators. The Allure Bipe has tips on all control surfaces which makes life much easier during setup and trimming. Not all designs have this feature.
Another thing noted was that the ailerons seemed quite hot. The Aileron rate was wound back a bit to alleviate this issue. The elevators felt nice and linear and soft which is not a lot different to the Alchemy. Bryan warned me about the rudder power of this model. I fly with a full rate rudder coupled with high expo (70%). The rudder felt quite good to me and certainly not over powered or savage by any means. It felt very linear to me. Perhaps I confused powerful with savage. They are quite different which is hard to explain here in words.
Ok, so now we have some observations. I find it helpful NOT to go making wholesale changes all at once. I normally pick on one issue at a time and take it from there. It might be a bit slower but I get there in the end. Taking your time also helps to avoid overstepping the optimum settings.
I decided to pick on the rudder trim first as that should be relatively easy to resolve. Well, so I thought.. Thinking Logically, I thought that a little less right thrust was needed. Bryan’s designs feature very little right thrust so I’m not going to be able to reduce it much before it’s pointing straight ahead. A beneficial feature of the Adverrun Single Drive is the ability to easily make thrust adjustments and also re-space the spinner to nose ring gap if needed. Some selected M4 stainless steel flat washers were used to adjust the thrust line to achieve less right thrust. Washer thickness can vary significantly so I measured them and selected which one went where. Off to the flying field again for more testing / trimming.
It would be super to report that all was fixed with the rudder trim. Can’t lie though... It didn’t seem to change the rudder trim at all. Ok let’s give even less right thrust. Different thickness washers were fitted to achieve this. Out to the flight line again full of hope and optimism... Ah bugger! Negligible change again. What’s going on here? We’ve made two changes to the right thrust and haven’t seen any notable change in the rudder trim the model is carrying. Is there a greater power at play here? Has the penny dropped yet? ;-) It dropped with a thud for me. If you read some of the notes about Bryan’s trimming guide he thinks of the model like an arrow. Without the weight at the front the arrow would wander all over the shop. Similarly, but not the same, if the fletching or tail feathers are deformed, twisted or damaged the arrow will not fly straight. It’s also true that 90% of all trim issues are at the rear of the model. The knife edge observations noted earlier give some clue as to what the issue is...
During alignment and setup of the model a number of things were adjusted and set but one thing was missed. We firstly checked the elevators were straight with Bryan’s Precision Elevator Alignment sticks. Yes, that’s another use for the sticks. Both elevators were perfect. We also set the stab incidences to be exactly the same. The DXL360 allows you to get within 0.02º of perfection. What I didn’t check during alignment and setup was stab incidence root to tip. This has since been checked and as it turns out, the stabs do have some twist along their length in the positive direction but the left stab has 0.2º more than the right stab. Funny that. This would account for the thrust changes having little to no effect and also the very slight roll coupling in knife edge flight. This issue will be fixed soon to prove the theory but I expect it to fix the left rudder trim, fix the roll coupling and perhaps the very slight knife edge pull to the belly. Is this Triangualtion? ;-) I will report back soon on the results.
Some final thoughts - Athough it’s very early days with the Allure Bipe, I can already tell that the design is very well sorted. Off the building board the trim is very close to where it needs to be. Bryan’s guidance on where to start with the settings was bang on and helped immensely. I can certainly see the advantage of a Bipe over a mono particularly in the knife edge stuff. In knife edge the model seems to carve up the sky. I was also pleasantly surprised with how the models weight came out. This isn’t a small model by any means and is full Composite and painted. My final weight was within 20 grams of the projected weight which I find amazing. There’s usually something you forget to account for in weight estimates. Interestingly, I can’t honestly say that I miss the Contra with this model. The Bipe seems to have enough inherent drag for down line breaking not be a problem. In fact, I’ve found that more throttle stick is required to keep the model moving at a constant. This could be simply prop selection or more time needed on the model to get the throttle management balanced.
I look forward to continuing the trimming journey and learning P21... ;-)