Frazer Nash FN5 gun-turrets

[Fidd88]

Thanks, I have a little E-flight apprentice, as yet unflown, and in the process of joining the BMFA and a local club, as well as getting lots of preflight advice. I had hoped to get cracking on all of this in the Summer but CV19 put-paid to all that. I'm just using the RF9 for the procedural and consistancy point of view. After a career in aviation, I'm well aware of the difference when contending with a light airframe and fickle actual winds! No doubt I'll shortly be seeking advice as to how to glue bits of foam together!

[Fidd88]

Such news at there is: I'm now at the phase of preparing to start metalwork, to fabricate the curved geodetic channels required to build the fuselage, wings and tail, and know how I'm going to go about it. The geodetic channel will be extruded in it's required cross-section, with a wall-thickness of 0.8mm (31 thou"). When it comes to bending it, a steel cable with a heavy weight on it will cause it to bend as the extruded tube is bent around a wooden former by being "wound" onto it, the steel cable causing it to conform thereto.

Some 300 different formers being required just for the fuselage. These formers will be CNC cut, and arranged in pairs to create the required shapes when the path of the geodetic goes over the centre-line of the fuselage.

One of the chief problems to solve, has been not allowing the cross-section of the geodetic to collapse as it is bent. When Brooklands Museum were making new geodetics for R for Robert, they encountered the same problem, and my solution is similar, except for the steel cable. What will happen is I will have CNC cut, hundreds of alloy plates, each being a fairly snug fit inside of the cross-section, threaded onto the steel cable, with softer thin plastic washers between each pair of alloy inserts. This (hopefully!) will allow each geodetic channel to be progressively curved onto the former without deformation of the cross-section. That's the plan anyway.

The next issue was cutting the notches required in the channel, at each cross-over of geodetic members. This had me stumped for a long time, as the only thing I could think of was using a grinding-disc, and this ran a severe risk of creating tiny imperfections in the cut which would be a source of metal-fatigue. I eventually had the happy inspiration of using a cutting-disc on the angle-grinder, with a mount for the grinder and a "suitable system of gears and levers" to keep the depth of the notch, and the angles of the cuts precise and correct. This mount isn't designed yet, but it is buildable. The only remaining task is the horizontal cut at the top of the notch which has curved corners, lining up with the two angled cuts done with the cutting-disc. This will be done by hand with a purpose-made steel chisel, probably impelled hydraulically using an adapted bearing-press or jack.

Provided the preceding works, and it should, that's all the shaping problems sorted. The next thing to address is access of tools. The rivets I intend using are Avdel Micro-rivets, but the pneumatic rivet-gun supplied has a rather large nose, making getting it in where I need it to go, is a problem. Tomorrow I'm off to see them, to see if a custom nose can be made, or is viable for me to have made, and then to ensure the cross-section of geodetic allows such an adapted tool into where it needs to get. Once all that's squared away, the remaining tasks are to ensure a CNC company can fabricate thousands of the small fittings required to make the standard geodetic node bolt and rivet together. I attach a rather old picture of the two types of joint prevalent on the airframe. Which hopefully will allow the preceding to make sense!

So, nothing physically to shew for it, but lots of theoretical progress as to what's needed to go forwards. Should be an interesting year ahead!

[Wingco57]

Fantastic engineering, how do you keep the correct radius of the fuselage contour during bending? As they must be compound curves.

 

[Fidd88]

It's nothing but compound curves!  The formers are the shape of the path of the geodetic, not the fuselage as measured around say a bulkhead. But when installed, they will naturally create the shape of the fuselage. So one half of the formed shape covers the path as it goes upwards and backwards at 45 degrees from the longeron to which it is bolted, and then the other half describes the path from the midline backwards and downwards towards the longeron on the other side. The result should - If I've not bollocksed-up the theory, describe the shape of a geodetic passing from one side to the other, from forwards to aft as it describes the shape of the tapering fuselage.

This is the reason for the "measuring bars" on the jig. The first thing built will be temporary bulkheads which constrain long alloy tubular longerons the length of the fuselage. At very precise points, these have holes drilled in them to take the fitting (red in the picture above) which attach the beginning or end of a geodetic channel to a longeron. The measuring bars carry a sled with a plumb-line, and have measuring tapes on them, so I can simply slide the sled to the required distance, mark the longeron, drill it, and temporarily fit one end of the geodetic. When a crossing geodetic is fitted, I can then mark both where the notches need to be cut, and eventually take them all off and notch them before fitting everything for good. That may change in practice, but that's the plan.

The easiest way to figure this is to take an old football, draw some "latitude" (horizontal) lines in it, (these are the paths of the longerons), then lines that are great-circles (like the equator or any line of longitude) where the centre of the circle described is also the centre of the ball. The great-circle paths drawn are at 45 degrees to the latitude lines, and cross each other. Result - shape of football defined by both. Same principle, only a little more complicated, with the fuselage.

[Fidd88]

Well I'm just back from the rivet-gun shop (more of a factory really) and the absolutely brilliant news is that the tool will fit on the gaps I have to work within on the current dimensions. They even gave me a tool nose to take home so I can draw it up in CAD and play around with it. That means a lot less work potentially! - And building stuff sooner! Loud cheers!

[Fidd88]

On 10/22/2020 at 8:44 PM, Wingco57 said:

Fantastic engineering, how do you keep the correct radius of the fuselage contour during bending? As they must be compound curves.

 

I'll try and answer this again. Lets say our fuselage was cylindrical, and there are no longerons to complicate matters. If we cut through the cylinder at 45 degrees, we end up with the cut face of the solid-cylinder being an ellipse. So, if we make a solid flat ellipse of the same size, and wind a soft-metal shape around that ellipse, when the resultant elliptical alloy channel is fitted at the correct 45 degree angle, the internal space will be circular again as viewed from the end of the cylinder.

The same principle holds true if our cylinder tapers, only now the start-point of the channel must be the same as the point on irregular but roughly ellipsoidal (ovoid) path, in order for the taper to appear smooth and not irregular.

As the required former shapes (our ellipses in the example above) are easily derived from slicing a solid 3d shape, it follows that provided the initial 3d shape is correct, so will the shape of the eventual metal-basket of geodetics be.

As for actually bending the extruded channel. Inside the channel is a steel cable, of 2mm diameter, which has a breaking strain of circa 250kg. Threaded onto this are metal plates, and hard rubber plates, alternating in sequence, fractionally undersize to the internal cross-section of the extruded channel. On one end of our cable we attach to a 230kg weight on the floor. The other end attached a point on the wooden elliptical pattern laying flat. The pattern has an axle though the middle, and a ratchet mechanism, so that as the ellipse is rotated the extruded channel has to conform to the ellipse, via considerable forces exterted by the weight now being lifted off the floor, and transmitted to the channel via the alternating hard and soft plates and the steel cable. The result is dependant on the thinness and number of said plates, but should theoretically, create a geodetic channel of the correct curvature via hundreds of slight bends, but once the cable and inserts are removed, with a cross-section which is largely undeformed.

Phew.

There's a lot to do to get to that point, but it's the only way I can see of achieving it after 7 years studying the problem. If you have other ideas please let me know!

 

[Wingco57]

Nope, sounds like you cracked the code.

[Fidd88]

I'm unable to work these days - legacy of an old car-accident, so have decided to see if anyone wants one of these models, and is prepared to pay to have one built. To that end, I've put "commissioning one" onto Ebay to see if there's any interest. As they take 5-8 months to build and paint, and thousands of pounds to print all the parts to the required finish and dimensional accuracy, these are far from a cheap model to build. If I get a lot of interest I may look into getting parts injection-moulded, which would bring the per unit cost down, but raise the initial tooling costs.

The item number on Ebay is 184506409292

or here: Ebay auction to start commisioning process

I'm hoping, if I get a commission, to plough the funds back into the Wellington project, which fairly eats money! I should emphasise, this is not posted out of any expectation that anyone here would want one at this price!

[Fidd88]

Hi chaps. What with Covid, and it's effect on lots of companies, I'm struggling a bit to make progress at the moment. The rivet tool company I'm working with are having difficulties determining the tool-head to use to apply these very small (2mm diameter) pop-rivets. Until the tool-heads size is known, I can't do the CAD to finalise the extruded alloy profile or the sheer and gusset fittings, or the inserts needed to curve the extrusions.

So, instead, I've been having a clear-out and reorganisation in the workshop to gain more space to the right of the bench, which I'll need for operating the device I will build to curve the channel without deforming the cross-section.

Here's a film of changes to the workshop and my "patent" peg-board and shelving. which I find a really flexible storage system; and, a quick look at the two types of joints on the airframe.

 

[Fidd88]

Hi all,

I trust you're all keeping out of the path of this bloody virus? I was not surprised today when the British PM "cancelled Christmas" by doing away with a 5 day relaxation of the rules to enable us to get together. People just don't get it. We're in the fight of our lives against this damn bug, and being able to get together or go to the pub or whatever must come a very distant second to keeping the infected people requiring treatment below the point where our health services cannot cope - or else we'll cop for it like the Italians and New Yorkers did with mass-graves and much higher fatality rate.

I get so cross with people who lack the understanding or sense of duty that we simply must do everything possible to help each other, and help out doctors and nurses, until vaccinations can batter it back into some form of normality next year. My brother's a MD (Doctor) in the US, who has just had his first days off since March, having worked 7 days a week continuously in ICU since it hit. The stuff he tell's us about what he thinks about people who won't wear masks, or get vaccinated is blistering in it's raw anger, and he's not someone usually given to "anger".

Rant over. (apologies)

So, my daughter and I are spending Christmas alone, as we'd intended to even before it was "cancelled".

Regarding the Wellington: I'm still hung up on the rivet-tool decision, which makes progressing with the CAD or production of extruded alloy channel impossible. I'm dealing with the UK supplier of an American manufacturer of rivet tools, so everything takes twice as long. Looks like mid-January now before that log-jam can be broken. Instead I've just finished reorganising the workshop so that there's room for the "bending device" to operate on the right-hand end of the bench, and I'm also busy making lots of rough drawings of how the bending device will operate. Not quite there yet, but close now. The changes in the workshop also allow the jig to turn without colliding with the nuts and bolts storage.

Latest changes

 

[Fidd88]

Not much news, but at last a little progress. I've got a mate to turn up a rivet-gun tool-tip for placing the 2mm diameter rivets, with a long enough reach to place them in the majority of the "ordinary" geodetic joints. All 16 rivets and 2 machine-screws with nuts per single joint! I haven't counted all the joints on the Wellington, but I expect to use well over 20,000 rivets on the whole airframe.

I'm now redrawing all the joint fixtures and fittings, and the profile of the channel, as well as the inserts used to prevent the cross-section collapsing as it is curved. Have now decided to use small o-rings, or possibly small rubber washers, to place between the 1mm thick metal inserts which will exert forces from the steel cable on the geodetic channel to effect the curving. Pics to follow in due course when there's something to show.

This whole rivet-tool-tip problem has taken months to resolve. I'm now facing a series of things that need sorting out in a very particular sequence, so I commit the least money possible in each phase until I'm sure the whole thing is feasible.

[Fidd88]

After a couple of days at the pc the two main joint-types for the alloy "basket" which forms the fuselage and wings of a Wellington have been drawn up to ensure I can get all the internal metal fittings made before the extrusion die tooling costs are paid. 

 

[Fidd88]

Hi all,

More news on the Wellington project - but not "good" news. I'd budgeted around £1000 to have the two types of solid internal fittings for the geodetic joints, as shewn in the film above, and found a company in China who were able to undertake the work. When the quote came back it was (for 13,000 fittings) over £17,000! Which is a complete show-stopper if I can't work around it.

I'm currently investigating making the "butterfly" and "wishbone" joints by folding flat sheet cut to shape - which is a much cheaper proposition to have done, but may be problemtical in practice, and may do the job. Both are also possible to make from an extrusion, milling the shapes one at a time, to create the finished parts. Needless to say, this is not my preferred method as extrusion die tooling costs are not trivial. In the meantime I'll be seeking more quotes in case the 17k one was an outlier.

Not a happy bunny.

 

[Peterpools]

Mind blowing work for sure

Keep 'em comin

Peter

[Fidd88]

Some better news. I've had some assistance from some of the other RC chaps who've suggested using CNC to make a small number of these fixings, and to use them to make a mould from which wax positives can be cast and attached on a suitable wax sprue. These can then be investment cast in alloy, in numbers, and machined/drilled/ground as required after the casting. Which puts it financially at least, back on track, and I get to learn some more skills! I've also had the offer of a loan of a small forge, which is both kind and hugely good for morale! I'll also be looking into a different form of casting on a commercial basis, to see if a better result can be achieved within an affordable price.

[Jeff]

Well, that sounds encouraging at least, I was hoping you would be able to access some sort of result, as this is a VERY cool and ambitious project, that should be able to continue to fruition if at all possible, within financial reason of course...... most interesting Fidd, and looking forward to more.....

[Peterpools]

Always a pleasure to follow unique and interesting projects

Keep em comin

Peter

 

[Fidd88]

Hi lads, I can't tell you how brilliant it has been to get some encouragement at this juncture in the project, as well as going through a divorce after 26 years - the Wellington seemed stymied at every turn, and has been really hard to keep up the enthusiasm for it in the face of difficulties with the riveting, and then insane quotes for things I'd assumed - my error - would not be overly dear. I really needed some encouragement to get past this point, as the difficulties expenses and technical problems were seemingly stacking-up faster than I could deal with 'em.

So THANKS! Much appreciated!

[Maybach_Man]

Hi Tim..

Been following this project since i first saw the turrets on the other forum. Sadly i couldn't comment over there as i'm not a subscriber.....but by happenstance came across you on this forum...so all good. Great work so far. You have probably forgotten more about engineering than i will ever know... Two things.... could you get a milling cutter ground to machine the node cut outs two at a time (back to back) instead of a grinding jig?..  Also, have you had a look at continuous guttering forming. These guys have a machine that they carry on the back of a van and form guttering out of aluminium from a roll. Obviously the machine would be unsuitable...but perhaps the process might bear thinking about?... Anyway, enough of me blabbering on about things i plainly know nothing about...Keep Safe

[Fidd88]

Cheers! The issue with the rolling-machine making the channel is not difficult about the cross-section, it's curving the geodetic during that's the complex bit. Hence my decision to extrude first and curve after. Which, incidentally, is how the missing geodetics on R for Robert at Brooklands, was made.

I'm not a trained engineer at all, I worked in a motorcycle workshop years ago, rebuilding engines and servicing, so I've a little mechanical common-sense, and an O level in technical drawing from 40 years ago - indeed initial drawings for the turrets were pencil on paper at a drawing board! I just pick-up new skills from youtube, and if I collide with an immovable obstacle, just ask for help! Professionally, I was a commercial flying Instructor, teaching lads to get through their GFT's on light aircraft, which gives some familiarity with the flying side of the project, although 20 years after the car-crash that terminated my flying career, I've forgotten a LOT!

You're correct that a side-mill might do the job, but without trying it, I'm uncertain of results. I'll go with whatever is fast, tolerably accurate and endlessly repeatable. A CNC mill is actually, in my experience a Royal PITA, as they'll decline to work as you might wish, out of sheer bloody spite! 

[Wingco57]

IIRC Brooklands used shaped blocks that fitted inside the channel sections and bent them manually slowly into the correct curves

[Fidd88]

I know! I spent some time talking to some of the chaps that did it, with a single hundred plus year old fly-press -  and some determination! Talk about patience! You could have bottled my expression and sold for thruppence a time, once the full horror of building it thus sank into my conscious! Blimey, but that was a labour of love! My planned method looks positively Henry Fordesque compared with that - tiny movments at a time, and then drifting out the wood blocks afterwards. Crikey! The found a better frame 50 bulkhead (the one to which the trailing edge of the wing and aft end of the bomb beam attaches to) that some old boy donated to them - he'd been using it since collecting it in a scrap-yard at the end of the war - in his home-made greenhouse! Brooklands bought him a new greenhouse!

[Fidd88]

So, some good news now. I've finalised the design of the geodetic channel, and now know how the internal fittings can be made. Hope this week to start the process of getting the alloy extrusions made. Once they're in, I can check the actual dimensional accuracy of them, and if needbe, change the dimensions of the CNC'd parts from which moulds will be made prior to casting. Then the CNC'd parts can be made, and the mould-making can commence, from which wax patterns can be cast. After that they can be investment cast to produce the required parts in alloy. They can then be drilled with the required holes, and fettled to fit into the channel.

The next phase after that is making some small test-pieces, which I hope to get to by the end of the year. In parallel to this I'll be designing, building and testing the device for curving the channel to a specified curvature, and also a device for cutting the notches in each geodetic where one crosses another. This will be a 2 part operation, with two parallel cuts cut with an angle-grinder (thin cutting disc) in a jig restraining it's movement to achieve two cuts of exact depth and seperation. A special shaped tool (chisel) will then be used to press down on the channel to make the tapering portion of each cut. I'll probably acquire a small bench-top bearing-press, with some means of attaching the tool to the moving portion.

The other thing under consideration at the moment, is to make my Mk IC Wellington an early example, still fitted with the unlamented FN25 "mid-under" (ventral) "dust-bin" turret. I may prepare drawings for it now, but a decision on building that may depend on funds, and the actual C of G position when the airframe nears completion, as it's well aft of the C of G, and these model turrets are not light.

So, that's news for now!

[Fidd88]

On 9/25/2020 at 11:37 AM, Bomber_County said:

Having flown RC by all means use the PC version but I recommend the real thing as well. Find a club and become certified, every club have training sessions with buddy box controls, so if you get into difficulties they can help you out. You can buy an ARTF (almost ready to fly) cheaply, then add a transmitter.........

Last week was fun - I did my first flights with an RC aircraft, my little e-flite "Apprentice". Managed 5 flights, mainly circuit bashing, with no buddy-box, but a more experienced pilot next to me with timely advice. He gibbered a little, winced only a bit and hardly sobbed at all. I clipped a cow-slip in the undershoot on the 1st landing which span it into the long grass - no damage - and then pulled off 4 good landings thereafter. I had been practicing on the PC beforehand! I've since been flying a few times more, and am getting some consistency to my "circuits".

Learned lots, had a complete ball, and avoided the "walk of shame" with a bin-liner of broken foam! (Loud cheers)

I've also made a little progress on the Wellington front. My daughter's boyfriend is an apprentice-engineer like my daughter, and so I'd asked him to turn up a fitting for my rivet-gun suitable for these tiny 2mm rivets, which he supplied via #1 daughter today. It works beautifully, and I placed 5 rivets in a 1/2" sided square - like the dots on a die - through 2x 1mm plate alloy scrap. This enabled me to rotate the two sheets relative to each other to "work" the rivets to see if they'd fail. With a 9" mechanical advantage, and moderate force, I was able to loosen them, but it was something of an unfair test, as the mechanical advantage was considerable, acting on only 5 rivets. So I think they'll do. I've measured the witness-marks on the mandrels (the disposable shaft that snaps when you place a rivet), so can now get a 2nd tool made to place the rivets as deep as possible close to a perpendicular surface. This length of tool is a critical value, as it will greatly affect some design parameters. So, useful stuff occurring. I'm going to have to find a better way of de-burring after drilling, so that I don't eat away too  much metal. A counter-sinking bit does a good job, but takes away fully half the metal on 1 sheet!

:-(

 

 

[Peterpools]

Way over my head but amazing just the same

Keep 'em comin

Peter