[Canberrauav] Sunday Recap and future plans Stephen Dade stephen_dade at hotmail.com Sun Sep 2 22:04:02 EST 2012 Hi all, Today's testing was spent over on a farm near Sutton, where we performed our first complete auto flight with the Mugin (minus the manual landing). We also tested the complete setup of our ground station. There are a number of notes we made from this: -Add airspeed sensor calibration and testing to checklist -The horizontal antenna for the 900 MHz radio came loose (fixed with sticky tape, but we need a more permanent solution) -The heartbeat pin from the APM to the failsafe kept coming loose (broken?) and triggering the failsafe. This should be solved by the V2 failsafe being installed. -With the engine on, communication between the pilot and ground station was difficult. This was solved by using Andrew Moss as a go-between the two. -The airspeed sensor is still giving us bad data, perhaps due to bad placement. We will look at placing it in the nose or wing. In the meantime, the GPS groundspeed worked as an OK replacement. -The Bullet (5.8 GHz) on the ground station failed during flight. This was most likely due to network issues because it worked fine once it had been plugged into Tridge's laptop. It is possible that the network switch didn't like our inverter and failed. -The bottle (accidently) dropped just after the engine started. The APM definitely sent the drop signal to the servo, but we're not sure why. We could perhaps have an "arming pin" on the bottle drop mechanism this is removed just before takeoff. -We need to remember to point the ground station antenna towards to UAV -It was quite windy at the airfield. We were having some trouble keeping the ground station from blowing away. This could be solved by marking out (and then taping) particular items (laptops, etc) to the table. -We were quite distracted/panicked by the various issues we encountered during the day, leading to a lack of briefing, monitoring and communication. -We need to ensure all items are clear of the prop once the engine has started. We could put some bright tape on the tail booms of the Mugin advertising this. -Start pickets are easier for the Mugin's holdback once the engine has started. This should be done with the Mugin just off the runway, facing away from it. This way, the Mugin can be easily pulled back onto the runway for takeoff. -We need a dedicated safety officer, watching for any potential issues. I believe Andrew Moss has volunteered for this. -The R/C failsafe needs to be added to the checklist -We need a sunshade for the ground station Bullet, so we can see it's status lights more easily -We also need mousepads. The table is not a friendly surface for mice. ---------------------------------------------------------------------------- So, where to from here? The plan for this coming weekend is to go through the ground station setup, procedures and jobs. We will setup the ground station at CMAC and do a few "missions" with one of the smaller UAV's. If it is available, we may go back to Sutton and do another run (weekend on the 16th/17th). The risk of this is that if the Mugin is lost/destroyed, we will not have enough time to get the Beaver up and running as a replacement. Also - I'd like to get copies on any photos/videos from today's testing. I'll be at MHV on Tuesday if I can grab them there. Thanks Stephen
[Canberrauav] Sunday Recap and future plans Andrew Tridgell tridge at samba.org Mon Sep 3 09:28:52 EST 2012 Hi Stephen, > -The horizontal antenna for the 900 MHz radio came loose (fixed with sticky > tape, but we need a more permanent solution) the horizontal antenna is for the 5.8GHz radio. The 900MHz radio has two vertical antennas. > -The heartbeat pin from the APM to the failsafe kept coming loose (broken?) > and triggering the failsafe. This should be solved by the V2 failsafe being > installed. It wasn't a physical problem - the fix was to reset the APM2. I suspect the output pin did not initialise correctly on startup. > -The Bullet (5.8 GHz) on the ground station failed during flight. This was > most likely due to network issues because it worked fine once it had been > plugged into Tridge's laptop. It is possible that the network switch didn't > like our inverter and failed. Note quite. The network cable or switch between my laptop and the bullet failed during pre-flight testing. I then connected the bullet directly to my laptop. That connection worked correctly all through the auto flight (I have checked the logged images on the GCS, and it has images that came over the 5.8GHz link for the whole flight). > -The bottle (accidently) dropped just after the engine started. The APM > definitely sent the drop signal to the servo, but we're not sure why. We > could perhaps have an "arming pin" on the bottle drop mechanism this is > removed just before takeoff. I've looked at this in the log, and the drop happened when the APM2 saw a change in value of the RC input channel 7 from 1300 to 1498. The RC7 input channel should not affect the RC 7 output channel, so this is an ArduPlane RC_Channel_aux bug. Randy tells me he fixed a bug that might cause this yesterday, but I will be looking closely at the code to see if the fix could have explained the behaviour we saw. > -The R/C failsafe needs to be added to the checklist yep. Key items before the "record failsafe position" button is pressed on the receiver: - sticks and trims must be in correct positions - ignition must be on - mode must be AUTO the pre-flight check should be to switch off the transmitter with the engine off. The pilot should check that the ignition light comes on, and the GCS operator needs to run "status RC_CHANNELS_RAW" and check all channel values are correct. > If it is available, we may go back to Sutton and do another run (weekend on > the 16th/17th). The risk of this is that if the Mugin is lost/destroyed, we > will not have enough time to get the Beaver up and running as a replacement. do you want me to ask Ben if we can do this? Cheers, Tridge
[Canberrauav] Sunday Recap and future plans Chris Gough christopher.d.gough at gmail.com Mon Sep 3 12:57:25 EST 2012 I have another item to add to the list from yesterday: wing spar fatigue During post-flight inspection, Jack noticed an apparent seam in the wing spar (joiner tube), almost as though it had been joined with a weld. Closer inspection found there were two of these seams, one coresponding to each wing/fuselage junction. These seams are clearly visible on the inside as well, looking down the joiner tube in bright sunlight. There is also a small dint in the tube about a centermeter inside one of the seams. Neither Jack nor I had noticed these before. These marks could just be caused by rubbing, but I think they could also be clues of immenent catestrophic failure. The fact they print through to the inside of the tube has been weighing on my mind all night. I think it's also possible the dint is a crinkle (first stage of a fold). We ordered a replacement wing joiner tube but it didn't arrive. The spar is slightly bowed and appears that the airframe was built arround the bow - it's unlikely a replacement would be (or could be) bent in exactly the same way. Because of this slight bow, the orientation of the joiner tube during assembly affects how the wings sit against the fuselage. Yesterday, there was a few mm gap between each wing root and the fuselage, and I think in that condition the stress on the joiner tube is _much_ higher than when the wing roots are hard pressed against the walls of the fuselage. In the hard-pressed-against-the-fuselage situation, the joiner tube is mostly in strain because the compressive forces are transfered to the fuselage through the entire wing root. When there's an air gap, the joiner tube has all the same strain plus 100% of the compression force is concentrated in a few square cm aluminum (i.e. in approximately the zone where the dint appeared!). I think we should consider stout bracing on the inside of the tube, expecially to add compressive strength. I have come up with one option based on materials I have to hand, I think we should discuss it (and maybe others) tomorrow night. Jack, can you please bring the joiner tube to MHV. The idea I have so far is to start with a 700 sized helicopter boom. It's about 1m long and very stiff, but not strong enough on it's own to do the joiner tube's job. It's thicker than a large thumb, but quite a bit thinner than the joiner tube - expect it would be a very loose fit inside. My idea was to cut off a suitable length of boom, then built it up it's diameter by wrapping it in woven fiberglass sheet until it's a tight/snug fit inside the joiner tube (dry, before epoxy) - It would probably take 2-4 mm of fiberglass to get it nice and tight. I could then wet it out with epoxy and press it into position with some kind of plunger. I guess it might make the joiner tube a few hundred grams heavier, but wouldn't affect the ballance (because the joiner tube is right on CG). Also, our assembly procedure should try to ensure that there is no gap between the wing and the fuselage. Epoxy cures fully in 48 hours, in my heated fume cupboard it's almost fully cured in ~12 hours (I call it a fume cupboard because it smells, not because it's got an extractor :). If we make a decision on Tuesday night, I'd have a couple of days to act and we could still fly with it on the weekend. Chris Gough On Sun, Sep 2, 2012 at 10:04 PM, Stephen Dadewrote: > Hi all, > > > > Today’s testing was spent over on a farm near Sutton, where we performed our > first complete auto flight with the Mugin (minus the manual landing). We > also tested the complete setup of our ground station. > > > > There are a number of notes we made from this: > > -Add airspeed sensor calibration and testing to checklist > > > > -The horizontal antenna for the 900 MHz radio came loose (fixed with sticky > tape, but we need a more permanent solution) > > > > -The heartbeat pin from the APM to the failsafe kept coming loose (broken?) > and triggering the failsafe. This should be solved by the V2 failsafe being > installed. > > > > -With the engine on, communication between the pilot and ground station was > difficult. This was solved by using Andrew Moss as a go-between the two. > > > > -The airspeed sensor is still giving us bad data, perhaps due to bad > placement. We will look at placing it in the nose or wing. In the meantime, > the GPS groundspeed worked as an OK replacement. > > > > -The Bullet (5.8 GHz) on the ground station failed during flight. This was > most likely due to network issues because it worked fine once it had been > plugged into Tridge’s laptop. It is possible that the network switch didn’t > like our inverter and failed. > > > > -The bottle (accidently) dropped just after the engine started. The APM > definitely sent the drop signal to the servo, but we’re not sure why. We > could perhaps have an “arming pin” on the bottle drop mechanism this is > removed just before takeoff. > > > > -We need to remember to point the ground station antenna towards to UAV > > > > -It was quite windy at the airfield. We were having some trouble keeping the > ground station from blowing away. This could be solved by marking out (and > then taping) particular items (laptops, etc) to the table. > > > > -We were quite distracted/panicked by the various issues we encountered > during the day, leading to a lack of briefing, monitoring and communication. > > > > -We need to ensure all items are clear of the prop once the engine has > started. We could put some bright tape on the tail booms of the Mugin > advertising this. > > > > -Start pickets are easier for the Mugin’s holdback once the engine has > started. This should be done with the Mugin just off the runway, facing away > from it. This way, the Mugin can be easily pulled back onto the runway for > takeoff. > > > > -We need a dedicated safety officer, watching for any potential issues. I > believe Andrew Moss has volunteered for this. > > > > -The R/C failsafe needs to be added to the checklist > > > > -We need a sunshade for the ground station Bullet, so we can see it’s status > lights more easily > > > > -We also need mousepads. The table is not a friendly surface for mice. > > > > > > ---------------------------------------------------------------------------- > > So, where to from here? > > > > The plan for this coming weekend is to go through the ground station setup, > procedures and jobs. We will setup the ground station at CMAC and do a few > “missions” with one of the smaller UAV’s. > > > > If it is available, we may go back to Sutton and do another run (weekend on > the 16th/17th). The risk of this is that if the Mugin is lost/destroyed, we > will not have enough time to get the Beaver up and running as a replacement. > > > > > > Also – I’d like to get copies on any photos/videos from today’s testing. > I’ll be at MHV on Tuesday if I can grab them there. > > > > > > Thanks > > > > Stephen > > > > > _______________________________________________ > Canberrauav mailing list > Canberrauav at canberrauav.com > http://www.canberrauav.com/mailman/listinfo/canberrauav > -- .
[Canberrauav] Sunday Recap and future plans Jack Pittar jpittar at bigpond.net.au Tue Sep 4 07:53:35 EST 2012 Hi Chris, I expect the dint and maybe some of the bending is occurring as a result of the sharp deceleration at the end of the flight in the grass. I still think strengthening the spar is a good idea, but using fibreglass resin inside the tube might be a bit difficult to get right, and then impossible to re-do if it isn't right. I suggest an alternaitive would be to obtain an aluminium tube that fits tightly inside the spar and tap / hammer it into place. We may have to machine a section of tube down to obtain this. Anyway, I'll bring to MHV tonight to discuss it. Jack. -----Original Message----- From: canberrauav-bounces at canberrauav.com [mailto:canberrauav-bounces at canberrauav.com]On Behalf Of Chris Gough Sent: Monday, 3 September 2012 12:57 PM To: Stephen Dade Cc: canberrauav at canberrauav.com Subject: Re: [Canberrauav] Sunday Recap and future plans I have another item to add to the list from yesterday: wing spar fatigue During post-flight inspection, Jack noticed an apparent seam in the wing spar (joiner tube), almost as though it had been joined with a weld. Closer inspection found there were two of these seams, one coresponding to each wing/fuselage junction. These seams are clearly visible on the inside as well, looking down the joiner tube in bright sunlight. There is also a small dint in the tube about a centermeter inside one of the seams. Neither Jack nor I had noticed these before. These marks could just be caused by rubbing, but I think they could also be clues of immenent catestrophic failure. The fact they print through to the inside of the tube has been weighing on my mind all night. I think it's also possible the dint is a crinkle (first stage of a fold). We ordered a replacement wing joiner tube but it didn't arrive. The spar is slightly bowed and appears that the airframe was built arround the bow - it's unlikely a replacement would be (or could be) bent in exactly the same way. Because of this slight bow, the orientation of the joiner tube during assembly affects how the wings sit against the fuselage. Yesterday, there was a few mm gap between each wing root and the fuselage, and I think in that condition the stress on the joiner tube is _much_ higher than when the wing roots are hard pressed against the walls of the fuselage. In the hard-pressed-against-the-fuselage situation, the joiner tube is mostly in strain because the compressive forces are transfered to the fuselage through the entire wing root. When there's an air gap, the joiner tube has all the same strain plus 100% of the compression force is concentrated in a few square cm aluminum (i.e. in approximately the zone where the dint appeared!). I think we should consider stout bracing on the inside of the tube, expecially to add compressive strength. I have come up with one option based on materials I have to hand, I think we should discuss it (and maybe others) tomorrow night. Jack, can you please bring the joiner tube to MHV. The idea I have so far is to start with a 700 sized helicopter boom. It's about 1m long and very stiff, but not strong enough on it's own to do the joiner tube's job. It's thicker than a large thumb, but quite a bit thinner than the joiner tube - expect it would be a very loose fit inside. My idea was to cut off a suitable length of boom, then built it up it's diameter by wrapping it in woven fiberglass sheet until it's a tight/snug fit inside the joiner tube (dry, before epoxy) - It would probably take 2-4 mm of fiberglass to get it nice and tight. I could then wet it out with epoxy and press it into position with some kind of plunger. I guess it might make the joiner tube a few hundred grams heavier, but wouldn't affect the ballance (because the joiner tube is right on CG). Also, our assembly procedure should try to ensure that there is no gap between the wing and the fuselage. Epoxy cures fully in 48 hours, in my heated fume cupboard it's almost fully cured in ~12 hours (I call it a fume cupboard because it smells, not because it's got an extractor :). If we make a decision on Tuesday night, I'd have a couple of days to act and we could still fly with it on the weekend. Chris Gough On Sun, Sep 2, 2012 at 10:04 PM, Stephen Dadewrote: > Hi all, > > > > Today’s testing was spent over on a farm near Sutton, where we performed our > first complete auto flight with the Mugin (minus the manual landing). We > also tested the complete setup of our ground station. > > > > There are a number of notes we made from this: > > -Add airspeed sensor calibration and testing to checklist > > > > -The horizontal antenna for the 900 MHz radio came loose (fixed with sticky > tape, but we need a more permanent solution) > > > > -The heartbeat pin from the APM to the failsafe kept coming loose (broken?) > and triggering the failsafe. This should be solved by the V2 failsafe being > installed. > > > > -With the engine on, communication between the pilot and ground station was > difficult. This was solved by using Andrew Moss as a go-between the two. > > > > -The airspeed sensor is still giving us bad data, perhaps due to bad > placement. We will look at placing it in the nose or wing. In the meantime, > the GPS groundspeed worked as an OK replacement. > > > > -The Bullet (5.8 GHz) on the ground station failed during flight. This was > most likely due to network issues because it worked fine once it had been > plugged into Tridge’s laptop. It is possible that the network switch didn’ t > like our inverter and failed. > > > > -The bottle (accidently) dropped just after the engine started. The APM > definitely sent the drop signal to the servo, but we’re not sure why. We > could perhaps have an “arming pin” on the bottle drop mechanism this is > removed just before takeoff. > > > > -We need to remember to point the ground station antenna towards to UAV > > > > -It was quite windy at the airfield. We were having some trouble keeping the > ground station from blowing away. This could be solved by marking out (and > then taping) particular items (laptops, etc) to the table. > > > > -We were quite distracted/panicked by the various issues we encountered > during the day, leading to a lack of briefing, monitoring and communication. > > > > -We need to ensure all items are clear of the prop once the engine has > started. We could put some bright tape on the tail booms of the Mugin > advertising this. > > > > -Start pickets are easier for the Mugin’s holdback once the engine has > started. This should be done with the Mugin just off the runway, facing away > from it. This way, the Mugin can be easily pulled back onto the runway for > takeoff. > > > > -We need a dedicated safety officer, watching for any potential issues. I > believe Andrew Moss has volunteered for this. > > > > -The R/C failsafe needs to be added to the checklist > > > > -We need a sunshade for the ground station Bullet, so we can see it’s status > lights more easily > > > > -We also need mousepads. The table is not a friendly surface for mice. > > > > > > -------------------------------------------------------------------------- -- > > So, where to from here? > > > > The plan for this coming weekend is to go through the ground station setup, > procedures and jobs. We will setup the ground station at CMAC and do a few > “missions” with one of the smaller UAV’s. > > > > If it is available, we may go back to Sutton and do another run (weekend on > the 16th/17th). The risk of this is that if the Mugin is lost/destroyed, we > will not have enough time to get the Beaver up and running as a replacement. > > > > > > Also – I’d like to get copies on any photos/videos from today’s testing. > I’ll be at MHV on Tuesday if I can grab them there. > > > > > > Thanks > > > > Stephen > > > > > _______________________________________________ > Canberrauav mailing list > Canberrauav at canberrauav.com > http://www.canberrauav.com/mailman/listinfo/canberrauav > -- .
Copyright 2012 http://canberrauav.org.au/