Saturday, 21 December 2019

French Twister article from 2013

For some reason this article had slipped by me until recently. I bought a copy of the magazine and share the article here.

Whilst my French is non-existent I get the impression the author was very impressed with the Twister.

Sunday, 17 November 2019

Trim fail

I had a bit of a scare last week.

Just after take off I heard a 'bang' and had to use more back pressure on the stick than normal to keep the climb attitude the same.

I decided to continue to climb up to a safe altitude - 3,000 ft - before investigating what had gone wrong.

After thinking about it for a little while I concluded that one of the trim springs must have either broken or come loose.

The rest of the flight continued without incident and I made a normal landing. The stick forces were not much more than normal - thankfully the Twister does not really need a trim in the first place as the forces are so light.

That evening I investigated and sure enough the rear spring securing bolt had come right out. 

This explains the stick forward bias as only the front spring was acting against the elevator push rod tube - requiring a constant rearward force to overcome it.

I noticed that the bolt thread did not have evidence of Loctite on it so that is a failure on my part. It's taken just over 400 hours for this to happen.

I've used Loctite this time! And also added TorqueSeal to the outside so that at each annual it can be inspected for movement.

Something to be aware of if you have not used Loctite - but certainly an inspection item (with the TorqueSeal) to add to the Annual if you have already not done so.

After removing the baggage tray it was instantly obvious what had happened. The bolt holding the rear spring had worked loose and come completely out.

Closer view.

After fixing with Loctite on the thread of the bolts and also now with TorqueSeal (in Blue) which will aid inspection.

Saturday, 9 November 2019

New spinner

It has taken me many months but I have finally finished getting a new spinner installed after the old one 'departed company' in May.

In my usual fashion it has been 'modded'. I always found the original design of spinner to be a bit blunt looking compared to the rest of the aircraft, so this was my opportunity to change it's shape.

My boss loaned me an old spinner from his plane and I used that to make a 'splash mold' part off the front of it. I then combined that part with a replacement spinner that was kindly sent by Matthias.

The result is a sharper look - more Mark 9 Spitfire than the old rounded Mark 1 looking spinner.

The other modification is the addition of a front plate. This has added a great deal of rigidity and security to the attachment of the spinner, I feel quite confident it will not come off again!

The front plate is made using the same schedule as the back plate. A 3mm thick sheet of carbon is cut out and drilled to match the bolt holes in the crush plate. Then an angled flange of carbon is added to the edge, projecting forward, giving a large area of contact with the inner front of the spinner.

The spinner was fitted and aligned using a laser level before all the holes were drilled. Then nut plates were fitted to the inside of the front and back plate flanges. 12 on the back plate and 4 on the front plate. I've gone for metric anchor nuts this time round and button head Allen screws with fibre washers underneath for security as I found the Tinnermans with the sprung plate to be unreliable over time.

Everything runs very smoothly now. However I will still do a dynamic balance to see what the vibration levels are and if they can be improved at all.

Regarding the old spinner - the conclusion was that it was a substandard part. 

It was not made by the factory but supplied to me by Hercules propellors using their own mold. They had not followed the same layup schedule as the factory spinner. They only put carbon on the back inch or so where the fasteners go for the backplate - the factory made spinners have carbon all the way up to and beyond the cut outs for the propellor. Therefore the old spinner was vulnerable to cracking in the prop cut out area. 

I believe this is what must have happened, a crack formed and then quickly propagated on the fateful flight, the spinner ripping open once the high pressure air got inside and popped it off, wrenching the fasteners as it went. We will never know for sure exactly what happened as the spinner was disintegrated but the above seems to be the most likely cause. Please check your spinner if you believe it was made by Hercules as they are not fit for purpose.

Original and new shapes compared.
Front plate for added security, micro on the edge was done to form fit to the inside of the spinner.

Side view of front and back plates

All done and looking good

I think this shape of spinner suits the Twister better

Button headed Allen screws with fibre washers underneath with a front plate. This is the same standard used on certified aerobatic aircraft

Sunday, 20 October 2019

First take off of a D-Motor engined Twister

Great news from Germany as the first D-Motor engined Twister takes to the skies for the first time.

Video on this link:

Reports are that all engine parameters stay in the green and with a climb prop they are seeing 1,000 fpm climb rate at 75 knots and 100 knots in level flight at 2,800 rpm (max continuous).

Next test flights will be with a cruise prop.

Monday, 19 August 2019

400 hour service on UL Power engine

For the second 200 hr service I decided to visit the UK UL Power agents, Metal Seagulls, based at Sleap airfield.

Mostly because I wanted Jonathan Porter to look over the engine and also because I do not have the facility to do a compression test on the cylinders, which is called for as part of the 200hr scheduled service.

Jonathan is very knowledgeable and accommodating, readily sharing his wealth of knowledge in an open manner. We are really fortunate to have such a person as the agent for UL Power here in the UK.

We had also planned to look at the cylinder head gasket on number 3 as I had long suspected a slight leak from the head gasket and this certainly had evidence of it throughout its life with a 'spattering' of 'black stuff' (sticky by product of combustion) that continually leaked out and was spattered against the pushrod tubes and inner cowling since the very first 15 hour service up until now. As it turns out the 'black stuff' has reduced quite a lot recently and almost certainly has the effect over time of gumming up any leak to the point where it becomes a non-issue. So Jonathans conclusion was to leave it be.

The service went smoothly with only two issues.

The oil thermostat has failed again (stuck in the closed position so the oil cooler was not being used as part of the oil circuit). It last failed at 278 hours and was replaced with a new part. It failed this time after just another 100 hours at approx 380 hours. I noticed the oil temps again were not coming down as per usual in the cruise and stayed at 103 degrees. I noticed this about 10 hours ago. The oil thermostat has been replaced again (as a warranty item) and oil temps are back to their usual 90-92 degrees in the cruise which is optimal.

The second issue and a more major one is a leaking exhaust valve/seat on Cylinder number 3.

The compression test (done hot) gave these results:

Cylinder 1 - 78/80
Cylinder 2 - 78/80
Cylinder 3 - 69/80 (leaking exhaust valve/seat)
Cylinder 4 - 78/80

For comparison at 200 hrs they were all between 78 - 80 over 80.

An audible hissing could be heard from the exhaust when Cylinder 3 was being tested and putting your hand over the exhaust pipe you could feel and hear the difference, confirming that it was the exhaust valve/seat that was leaking.

This ties in with what Pete Wells was telling me recently - that the UL 260 cylinders are only good for approx 400 hours then they begin to fail. Especially at the exhaust valve/seat area.

So this is happening to me also despite me using the Decalin additive which helps to break down the lead in AvGas into a powder so it does not stick on the valve and damage the seat.

Lycomings and Continentals do not have this problem as they run much hotter CHT's and also they must have different metallurgy in the valve seats to cope with lead deposits. UL Power have made an aircraft engine which will not run reliably to TBO with AvGas - which is kind of ridiculous really...

There are many (most) places in the world where you have no option but to use AvGas so this is a gotcha for anyone expecting to run their aviation engine to TBO on aviation fuel.

Of the 400 hours that I have run this engine only approx 150 of those have been on AvGas (always using the Decalin additive). The rest on MoGas or UL91.

The other ongoing problem is the Throttle Position Sensor which has failed again. It first failed whilst I was in the USA at approx 120 hours. I replaced it with a new one when I got home at 188 hours. It failed again at 315 hours (approx 120 hours on from replacement - which seems to be the failure region/mode for this part).

This time I took apart the TPS and cleaned out all the Carbon dust from the contact face and also noticed that the ultra thin wires that make the contact were a little bent in places so straightened those out and bent the base of them slightly out so the contact was better. This worked fine for about another 70 hours or so until it started failing again. So I have pulled it apart and done the same procedure and it works ok now but I am looking to get rid of this and go for the Hall Sensor throttle option when I can afford the plane to be out of the air for more than a week. At the moment I am using it to commute to work so it would be a hassle to have it out of the air for so long. 

The problem is that the ECU has to be removed and sent to Belgium for new firmware to be updated so this means at least a week lost to posting their and back not to mention the time fitting the new throttle sensor type and some wiring changes to accommodate it. (Not to mention the approx £600 cost!)

I plan to do this job and also get the valve seat on Cylinder 3 fixed in the Winter sometime when I can afford to be out of the air for a week or two.

The fact that UL Power no longer use this old style TPS and only make engines with a contactless Hall type sensor for the throttle now says it all. This was always a sub-standard component for aircraft use. I know of many other UL Power users that have had their TPS fail as well so it is not something unique to my installation.

I am also now using the UL recommended Teflon additive to the oil for aerobatic engines. See pic. This is UL Powers answer to the 'mysterious' seizure of the engine by oil starvation on G-JINX that Chris Burkett was flying during a display at Abingdon in May of 2017. UL's thinking being that the Teflon will continue to lubricate in the event of oil starvation. I would of course prefer that they got to the bottom of why the oil was not getting around the engine and why it seized in the first place rather than ad a 'sticking plaster' solution such as this. It is not cheap either at £36.40 per bottle! This adds quite a bit to the oil change cost each 50 hours.

At Metal Seagulls for the 400 hour service.

Teflon additive for aerobatic engines.

Throttle Position Sensor - note damage to thin contact wires.

TPS inside. Carbon dust forms on the contact faces so I clean these each time I have it apart.

Saturday, 10 August 2019

National Aerobatic Competition

Last weekend I competed in the National Aerobatics Competition.

My first flight was not up to my best standard (scoring 76.5%) so I was only in 6th place after day one.

Day two went much better with the highest score for the day (scoring 82.7%)for the 1st UnKnown sequence which put me in 2nd place overall and just 0.02% behind the leader.

Day three did not go well for the 2nd UnKnown as I messed up the first of the quarter clovers and ended up getting a hard zero (scoring 70.5%) which put me back to 5th overall from 14 who competed.

Figures flown and videos of each sequence are below.

1st Sequence flown which is the Known for the year.

2nd sequence flown which was the first of two UnKnowns.

3rd sequence flown which was the second of two UnKnowns.

Saturday, 20 July 2019

Pics from Paderborn

Johan arrives at the home of Twisters. His first 'international' flight in his Twister.

Trying out the open top Twister for size.

Great view and a feeling of freedom. Would be good on a hot day.

Open top is interchangeable in 30 seconds with the proper canopy. 
Matthias returns from his test flight.

Ingo takes the open top Twister up for a spin.

Ingo getting airborne.

Taxiing out for our four ship formation.

I had the best seat in the house with views like this.

The prototype Twister. Now resides in the Deutsches Museum Flugwerft Schleissheim in the north of Munich.

Sunday, 14 July 2019

Twister 4 ship formation

We had a memorable flight of four Twisters, at the home of the Twister, Paderborn Haxterburg, on Sunday 7th July.

Flight lead, Matthias Strieker D-MTMH

Number 2, Ingo Zoyke D-MTMN

Number 3, Johan Teerlink OO-162

Number 4 and cameraship, Andy McKee G-FUUN

Can we make it 8 Twisters next year? Let's hope so.

Click on this link for a short video:

Monday, 3 June 2019

Energy management and controlling the monkey brain

So there I was, part way into my second sequence of an aerobatic competition, when something happened.

I was looking to the left as I was just about to pull up to the vertical for the next figure which was a stall turn, so my eyes were on the sight gauge and horizon to the left.

I was not looking straight ahead at the time of the incident.

My first thought was 'Bird Strike' as I'd seen lots of gull's circling in thermals during the climb and during the first sequence earlier in the day. Fenland is quite close to The Wash so the coast is nearby and many sea birds are sharing the air.

There was a change in sound - a notable buzzing noise coming from the front of the aircraft. I immediately reduced the throttle to just above idle. 

A quick glance at the engine instruments confirmed nothing untoward with the engine so the sound reinforced my first thoughts of a bird strike to the prop.

Although it could also be a loose piece of bodywork or fairing which was 'buzzing' in the airstream? I didn't know for sure but suspected the prop despite there being nothing different vibration-wise from normal operation.

A glide approach was then started back to runway 26 - which was the wind favouring runway of choice below. I was very fortunate as to where and when this incident happened BTW, high and fast (a high energy state) and right over an airfield with 2 runways.

A call was then made on the box frequency to the Chief Judge to let him know I had a problem and I was ending the sequence. He acknowledged and I immediately went to the airfield frequency which was blocked due to an incoming aircraft 5 minutes out making a long drawn out call. I took the time to eyeball my situation some more and visualise the approach to 26. Once the other traffic and the tower had stopped talking I made a Pan-Pan call, stating I had a problem and that my intention was an immediate forced landing on 26. The tower continued to talk back to me and also talked to me when on finals giving me wind information but I was mentally blocking out most of what he said as it was irrelevant at this stage and my brain needed to fully focus on the aviating from now on.

During the approach I verbalised to myself out loud "No go-arounds!". This acts as a sort of 'order' so your monkey brain does not take over and do something stupid. I often used this technique when I was Gliding - which if you have ever done in a competitive environment is just a series of decisions, the winning pilot usually having made the best decisions as flying skills don't count for nearly as much as good decisions in the results.

Coming onto the final approach and now lined up with runway 26 I again verbalised my situation and gave myself an order to correct. "Too high, too fast!" - "Side Slip!". So a side slip was engaged to lose the excess energy I had.

Once flap limiting speed was reached and the runway was certain to be 'made' I went for full flap. A final little side slip to lose that last bit of excess energy and then a reasonable landing was made with no further damage done.

Now comes the question - if you have an excess of energy then should you go for best glide, that is; fly slowly and extend your thinking time, or should you just aim to get back on the ground whilst things are in your favour - in this case an empty runway. More time would not have helped me in this case - if anything I have learned from this experience that taking excess energy down low is a good thing and a safety net - typically in an emergency you can always get rid of energy but not put it back. So if I were to do it again I would do the same thing.

Once I had reduced the throttle to just above idle I treated the rest of the flight as a glide approach - something I do often as practise anyway so I'm quite familiar with the 'picture', performance and sensations that go along with it.

Having said that - if for some reason I had to add some energy there was still the option to do that with the engine running but that would have been with some risk - still preferable to crashing into trees or water if that had been the other options though. ie: the risk was worth it if the other outcome was worse. As it was I had an excess of energy so did not need to ask anything of the powertrain after the incident.

I was asked "Why didn't you turn the engine off" by one pilot (the only one out of the hundred or so that have talked to me about the incident to ask such a question BTW). The answer to that is quite simple and 4 fold. 

Firstly there was nothing wrong with the engine so shutting it down would not have 'saved' it from further damage (anyway engines can be fixed, people not so easily. As a pilot you should always be concerned with doing the least dangerous action to yourself rather than the least damaging action to 'some-thing'). 

Secondly the prop had been through the most intense energy state at the time of the incident, which was - high speed, high rpm. Once the throttle was reduced to just above idle and the airspeed reduced somewhat then the chances of the prop failing anymore was reduced to the point of 'highly unlikely'. 

Finally and this is the main point, if I had shut down the engine then I would have had a completely unknown performance state to deal with that could not be altered and one that I had never practised (a dead stick) so this would have been very unfamiliar to me, which would have increased the workload and therefore the risk.

So the safest thing to do was to keep the engine running - albeit at just above idle as that is it's smoothest setting and provides the airframe with the least amount of drag - much less drag than a stopped prop would do.

It was only when I had taxied in and shutdown that the cause was confirmed - the prop stopped with the blade in front of me that suffered the most damage and was split. So it was confirmed as prop damage after all. At this point I suspected the spinner causing it and a friend coming up to shake my hand confirmed that with "the spinner is gone" (you cannot see the spinner from the cockpit of a Twister).

Luckily there was no other damage to the airframe as the spinner was disintegrated into thousands of tiny pieces which made there way either over or under the wings and tail. There is a very small scratch on the outside of the cowl but other than that I can find no other damage at all.

I had cleaned the spinner and prop before this flight too and did not notice any screws missing or anything else untoward so it is still somewhat of a mystery as to why the spinner failed in the first place. I would only add to this that the spinner was made by Hercules Propellers and was not a factory supplied item. It was made to a different schedule and was said to me (by a third party) to be   'not fit for purpose' (not their actual words, I am being nice here). I will do a separate post later on about the new and improved spinner I am working on as a replacement.

As usual I was filming the sequence from the tail looking forward so there is video of the whole event. Even though the damage itself happens in part of a frame - it happens that quick - from nothing untoward to prop damage in an instant, the video hopefully shows something someone else can learn from.

I edited the video to show what happened and how I dealt with it. I hope only that it informs others in some way. Check your spinners! And also practise your forced landings. If you found the 'Verbalising orders' thing helpful then think about employing it - one day it may help you make the right choices in a high pressure situation and not let your monkey brain take over.

To that final point there are many examples over the years of seemingly very experienced and capable pilots doing the 'wrong thing' in an emergency due to letting their monkey brain take over. The gentle pulling back on the stick 'I don't want to crash feeling after engine failure' (when it should be pushed forward) which results in a spin into terra firma.

Video on this link:

The leading edge urethane material stayed together - despite hitting the spinner at approx 600 mph and then the prop split along the next weakest point back which is where the urethane is attached to the wood.

There is a small notch in the urethane which shows how much force is going on as you can hit this material with an axe and it does nothing normally.
No other damage to the airframe, so no harm done except to my wallet and total score in the competition. Second place was on the cards again until this happened.

Wednesday, 8 May 2019

Twister hybrid - More details

Here are some more details regarding the Turbine GenSets currently on offer.

If anyone knows of any others please let me know in the comments section below.

First up the MiTRE from Delta Motorsports.

Their latest model has an output of 35kW with a weight of approx 50kg.

It is quite a compact unit as you will see from the below pics.

I would need double this output and if anything have a slight reduction in weight. Bear in mind that some of that 50kg is for emissions reduction for the car industry which is not necessary in experimental aviation, so with that equipment removed the weight should go down. Also more exotic materials can be used to reduce weight further, although that will drive the cost up.

The battery pack is a 'moveable ballast' so if needs be It could be relocated to another area of the airframe - perhaps under the safety cell or even split in two with part of it in the baggage area (as Pipistrel do with their electric model to retain the correct C of G).

The other turbine Generator on offer is the TurboTech from France.

Unlike the Delta MiTRE this unit is specifically designed for aviation.

They are quoting an output of 55kW with a weight of 50kg.

With 40kg of fuel (54 litres) on board this unit has the equivalent energy of 1000kg of Lithium-Ion batteries!

The size of the Turbotech unit means it will be a squeeze to fit it under the Twister cowling and to include the electric propulsion motor. So the battery pack and inverter would almost certainly have to be moved elsewhere in the airframe to make room for it.

From their Linkedin picture it seems like Turbotech are using a very similar idea (but in reverse) to the pre-cooler on the Sabre (single stage to orbit space engine). I believe the Sabre pre-cooler is made from micron-thin-walled Inconel.

More about the Sabre here:

More about TurboTech here:

And finally a link to the Delta MiTRE which is a few years old so talks about 17kW rather than their latest 35kW version:

Delta Motorsports MiTRE turbine genset.

Installed in the bottom of a car.

Compact size of the Delta MiTRE is clear from this video screenshot.

Turbotechs offering. Larger in size than the MiTRE but getting close to the figures I need for the Twister. They are quoting 15 litres per hour in the cruise - same as my current fuel consumption.

Turbotech's recuperator design seems very similar to the Sabre Hybrid rocket engine.

Monday, 6 May 2019

Twister hybrid - revised version

It's been 7 years since my first design of a hybrid Twister. (See my previous March 2012 post).

A lot has changed in that time and I've also learned a lot more about the possibilities of electric hybrid powertrains.

For the immediate future a hybrid solution is the only one that works if you actually want to tour in your aircraft (which I do).

Fully electric is only good for aerobatics and local flying of no more than 1 hour.

So with my latest design the bias is with much less batteries and a normal fuel load.

Delta Motorsports here in the UK are developing a gas turbine generator as a sustainer/range extender for Ariel's new HiperCar and for Morgans new electric model, both coming out in 2020.

Normally a gas turbine is very inefficient - typically 15%. However through the use of a recuperator and some other secret tech Delta have managed to get the efficiency of their turbine up to 30%. This makes it a game changer.

Now the generator is as efficient as a conventional internal combustion engine.

The electric propulsion motor is 96% efficient so there are hardly any losses after the GenSet.

In terms of Aerobatic performance this new hybrid Twister would have the potential to compete at the Advanced level (although the roll rate would also have to be increased, the easiest way to do this would be to dispose of the flaps and go for full span ailerons although this would lengthen the landing distance by quite some margin).

Comparing to the Extra 200 and 300, which are both quite a bit heavier, the power to weight ratio is very much in favour of the Twister. With the Extra 200 being 540kg empty, plus 120 for fuel and pilot gives a power to weight of 0.22 kW per kg. The Extra 300 fares only a little better being 682kg empty, plus 120 for fuel and pilot the power to weight ratio is 0.28 kW per kg. The Hybrid Twister with 150kW and an estimated empty weight of 330kg (+ Pilot and fuel 450kg) would have a power to weight ratio of 0.33 kW per kg. This would give it a phenomenal climb rate of at least 3,500 fpm.

With that kind of climb rate then the lower VNE becomes less of a problem as the display/sequence can be flown in a smaller space and at a slower speed as there is less need to 'wind up' to a high speed to gain energy before starting a vertical figure.

Watching the new GenPro recently on YouTube it has shown me that with a superior power to weight ratio you do not need to fly so fast and take up so much room in the box. 

This suits the Twister as it has a lower VNE and G limits than the Extras. (160kts vs 217kts and +6-4 vs +-10G)

As you can see below the turbine would be housed in it's own heat proofed section of the cowl. The exhaust can be made parallel to the airstream for a little extra thrust. The turbine would be fed either by a dedicated NACA duct on the side (shown) or taking part of the air from the LoPresti inlets at the front - if there is some spare capacity left over from the liquid cooling requirements.

Hopefully the existing cowling and oil cooler scoop can be used.

With the small battery pack you would have 2 mins of max power - 201hp. Then in the cruise the power output of the GenSet would slightly exceed the propulsion motor requirements and therefore it could be used to charge up the batteries.

There is also some nice redundancy here. If the turbine GenSet fails after take off then you still have approx 4 mins of battery power to get around an abbreviated circuit and back on the ground.

Cost is a big unknown at the moment as the turbine is not ready for market. This design would also need a new variable pitch prop to deal with the greater power.

The existing fuel system can be used and with 30% efficiency from the turbine it should mean that the range is not compromised. Jet A-1 is less than half the price of AvGas in the UK too (and if you can buy at the corporate rate then it is even better as it is approx 20% of the price of AvGas). 

See the below illustration for details and layout.