Frazer Nash FN5 gun-turrets

[FME erk]

RAF Station Grimsby - home to 142 Squadron  Oct 1941 -> Nov. 1942

I did study that in the Technical drawing class but I cant recall it that well.

I still do some basic drawings usually of old airfield buildings, history may not forgive us for ripping them all down and bulldozing the rest . . . 

Apologizes for the drift from the true topic 

[Fidd88]

Not a problem. A little off-topic variation makes for lots of interesting stuff! I am completely with you about the loss of this history. As an ex (civilian) flying instructor, my personal pet-hate is the way airfields are now being renamed to lend a false sense of proximity to the nearest - but yet not close - city. Manston becoming "Kent International Airport" and "Shoreham" being "Brighton Airport" (until they have a crash, when it's "Shoreham" again!), "Kidlington" becoming "Oxford Airport" and so on ad nauseam.

To me, these airfield names evoke the smell of 100 octane and the whistle of opened canvas gun-ports, the line of a Spit's wing, the tick of a cooling Merlin, and the incredible sacrifices and courage of the men and women at these places 80 years ago. To have some ignorant bloody self-serving councilman "rechristen" them to inflate his "profile" in his town makes me bloody cross.

As for developers building housing estates on them..... I cannot safely comment on a moderated forum and give full vent to my feelings on the subject!

[GusMac]

Definitely a generational thing with Blender. My son taught himself it and has done some pretty amazing with it including most of his Higher Art & Design project. 

He's now at Uni doing Computer Games Technology so it's just the way their minds work nowadays! Pencils are prehistoric for them!!

[Fidd88]

#1 daughter and I were invited to a
commemorative event at the Brooklands Museum in Surrey, to exhibit the now
largely complete turrrets. The event marked the 80th anniversary of the
"Battle of the Heligoland Bight" where 24 Wellingtons were despatched in
daylight on an armed reconnaissance raid of Jerry warships in the area off
Wilhelmshaven. Half the attacking force were lost on the way back, or crashed
in England. This fairly definitively spelt the death-knell of the notion
that the "bomber would always get through", despite new-fangled hydraulic
turrets, and the RAF then adopted night-bombing until early '45. ('44 in
France). It was also this experience which made us all think the B17
daylight bombing was going to be a very bloody business.

Anyway, whilst we were there we got interviewed by the chap who makes most
or all of the video for the museum, and he's just sent me the link:
https://vimeo.com/380993532

Rather better than my efforts!

Fidd

 

[Bomber_County]

Stunning build congrats.......

[Sir Desmond Glazebrook]

Lost for words. Astounding.

[FME erk]

I think I would be tempted to have it as a static model rather than flying about . . .  I would be absolutely horrified if it crashed . . . .

All the hours it has so far taken to produce a masterpiece of engineering in such a great scale, never to be done again I would say  . . .

Certified as 'Grounded' 

 

[Bomber_County]

Having been an RC flyer in the past, I’d agree. That was one reason I gave flying.......

[Fidd88]

Thanks lads. Although the turrets took the best part of 4 years to build, this included tons of research, learning CAD, and endless revisions of the initial design, followed by many many revisions and reprints of individual parts where needed. During this time I also spent a lot of time on initial drawings of the fuselage.

Now that the design is finalised, I estimate I could build, both turrets from scratch, in circa 4 months for the pair, which is not excessive when compared to the time expended on similarly sized RC models, whatever the construction technique. The same will apply to the airframe, it's the setting-up of the jig, paths of longerons and cutting the MDF patterns for curving each geodetic member that takes the time.

Once the actual geodetic fabrication commences, after prototype test pieces are made to prove the build-method, then I'm expecting the construction to be fairly rapid. It's the preparation that takes all the time. So were the aircraft lost, then again, finance permitting, I'd expect the airframe to be completely rebuildable within 2-3 years. If the design is proven safe, and reliable, then I'd be in a position to start building a second airframe , and/or turrets in fairly short order.

The original design-brief I set out was to be able to film with FPV cameras from within the two turrets, from the pilots position, and from a position roughly atop the aperture for the old ventral turret, ie able to view the length of the fuselage fore or aft, as well as through the long side windows. Secondly to replicate the geodetics in alloy at scale, and to have no visible servos or uncamouflaged electronics or wiring within the turrets or fuselage to detract from the fuselage interior view.

The ability to get good in-flight footage from within the fuselage really drove everything from the construction techniques to the turret complexities, and of course it was much more interesting to design and build these as fully functioning operable turrets, than the usual two sticks of black-painted balsa poked through a simple cupola shape in clear plastic...

 

 

[Sir Desmond Glazebrook]

Is this a commission build? Are you a full time professional modeller? I am trying to get my head around it. Your skills are incredible.

I can only look at so many photos before my head goes into contractions.

All I can do is watch in awe. Then wander off in shock. 

You, Sir, have amazing skills.

My most sincere respects. And a tip of my Akubra hat. 

You are an inspiration. 

Cheers!

[Fidd88]

Good Lord no, I'm just a chap in a shed, and havn't made a model otherwise for nearly 35 years, when I last made an Airfix kit! Most of the skills needed I picked up over the course of the last 4 years. If there's interest, I may look into seeing if it can be produced as the basis for a kit by Airfix or the like. It'd be wonderful to get commissions to build or just design this sort of thing though, but I think I need to improve the design to make it easier to build. The turrets are my "middle-aged crisis" or whatever they call it. Some buy Ferrari's, I built turrets, which some wag called "turrets syndrome"!

[Fidd88]

So things have come to a bit of a grinding halt of late as I investigated several means of fabricating the mould required to vac form the curved front cupola windows. I looked at CMC cut MDF/micro-balloons, but it was a great deal of work, and when costed out, getting the mould 3d printed in ABS via a Chinese service came out only slightly dearer than the MDF route, possibly cheaper once all the required resin and microballoons were taken into account. I'll need to treat the ABS with some liquid plastic or filler to bring it up to a highly polished and thermally stable finish for vac forming. I'll let you know how that goes. Below is a picture of the two-part mould, the red portion being the area I need, the pink the surround. The split is low down away from the red, so no seam lines. The two part approach was required for cost purposes. Underneath is a honeycomb structure which I'll fill with expanding foam, to help make the mould incompressible as it's vac-formed.

[Jeff]

Fidd, old chum................ my brain could NEVER figure this kind of stuff out............... for us mere mortals, this is VERY interesting..... always enjoy your updates...... amazing project...

[Fidd88]

I'm glad you enjoy it, but what don't you follow? 4 Years ago I was still doing drawings on a drawing-board with sharp pencils! Pretty much everything I'm doing now has been learned in the interim, the great majority in the first 18 months. Learning CAD, especially in Autodesk Fusion 360 is a doddle for most shapes. As I think I said upthread somewhere, the only area I needed to seek help was in constructing complex curvatures, ie shapes like the one above where it curves in two or more directions. 3d printing is a complete doddle although with experience you'd discover the limitations of different sizes of part, type of plastic and method of deposition.

This is nothing the great majority of you couldn't pick up in a few weeks of applied practice, and a few months to get reasonably efficient with it. In a way, I'm surprised the scratch-building modellers aren't making more use of it to produce parts for models where the aftermarket kit parts don't carry the part in the scale you want, or simply don't make them. CAD/3d printing is a seriously powerful tool.

This evening I came across a rather useful looking process for polishing 3d printed ABS called "Cold Acetone vapour polishing" which gets rid of the great majority of layer-lines inherent in 3d printed ABS. Once that's done, the only issue remaining is to prevent the hot Perspex - or PETG etc - from sticking to the ABS when the frame of hot plastic is vac-formed over the mould. As this develops I'll let you all know, as the principles hold true if you're making a 1:4 scale turret cupola, or a 1:72nd fighter canopy from scratch...

Attached is a picture of the underside of the two part (now separated) mould above. The hexagons continue to the underside of the shell, and I plan to fill the voids therein with expanding foam, to help make the relatively thin working part of the mould more or less incompressible when the hot plastic is vac-formed downwards onto the mould. That's the plan anyway!

The other wrinkle, if you'll pardon the phrase, is that although I've done the CAD drawings for the side and overhead window panels, and these are slightly complex curves, I've decided to try and fit flat sheet and bolt it all in place before heating it with a hair-dryer. If I'm lucky this will allow the sheet to develop the compound curves needed without the expense of a 2nd mould. Got to be worth a try!

[Fidd88]

Ok..... as "Blackadder" would say: "I have a cunning plan"!

The first problem, is that I learned to my utter dismay, that the melting point of ABS is lower than that of acylic sheet heated to vac-forming temperature. Doh!

So, the ABS shape above simply cannot be vac-formed over directly.

Instead I have to make a mould:

I'll start by securing the ABS shape to a board covered in clingfilm. The ABS shape is then painted with about 3/4 inch or more of silicone rubber with thixotropic additives, extending over the shape and 2 inches wider than the shape all around. When that has cured in situ,  dovetails are cut into the silicone around the flat periphery of the silicone. Then the silicone is covered with 'Plaster of Paris' to a depth of an inch and a half or so, and whilst still wet, the P' of P' is covered by the thickest possible layer of Plaster of Paris impregnated material - "Mod-roc".

The Plaster of Paris/Mod-roc lump demoulds leaving the silicone adhering to the ABS shape, which is then demoulded in turn before being placed back into the P' of P' section of the mould, the dovetails helping to register one with the other. The silicone is then cleaned carefully with plain water so as not to leach out the chemicals which prevent it adhering to the casting material. Finally, the silicone mould, now in it's hard surrounding P' of P' is filled with high-temperature-tolerant resin and bulking material, which cures in situ. Once cured, the mould again demoulds leaving the silicone attached to the casting, which can then be carefully removed leaving the pristine and dimensionally accurate plug over which hot acrylic can be safely vac-formed. 

Which should result in some nifty properly curved windows for the front of the turret, from roughly 60 degrees either side of straight ahead. The side and top window panels will be made in flat sheet, bolted in place and then softened with a hair-dryer to conform as required to the gentle compound curves required. With a little luck this will obviate the need for a 2nd mould. 

Pictures and films to follow as this gets under-weigh. Sorry I've not been updating much, but I've been busy figuring this out and learning the required techniques.

Pictures attached are those of the yet-to-arrive ABS shape. The 3rd picture shews the dovetails registering the silicone with the hard Plaster of Paris case, although in this picture it is a 2 part, rather than 1 part mould. Same principle however.

Edited February 1, 2020 by Fidd88
clarification

[Fidd88]

Hi again all.

Blimey using silicone is a mucky business! You know when you get a bit of wrinkled sellotape stuck to your finger and you can't get rid of it, and end up merely transferring it from hand to hand in utter frustration? Silicone is like that a thousand-fold. No matter your good intentions and resolution for working cleanly and efficiently, one ends up like a 5 year old left momentarily unsupervised in a toffee-factory! Good "clean" fun and very interesting.

So I've started by using a catalysed T28 silicone for the initial "skin" coat - equivalent to the gel-coat if fibreglassing - which had the consistency of syrup and was the very devil to keep from overlapping the clay dam I'd made. This was suitable after 2 hours for the "thixo'ed" coat, which is much stiffer, - more akin to boarding-school porridge! - and was moved up towards the top of the pattern from below, as it practically "stands-up" even before curing. In two hours I'll do the final coat, and in 5 should be able to trim off the surplus silicone and cut the dovetails. I'll then leave it a couple of days - maybe 3 - before commencing the "Mod-roc" and "Plaster of Paris" hard shell. I'll make this thick enough to embed some rawlplugs, for attaching feet to the mould for when it's later inverted for casting.

Films in the usual place, first two here if anyone is interested. As usual, if it all still goes wrong, I'll leave the films up in case it helps anyone else!

 

Initial work on mould

Fun with silicone - mould-making

[Fidd88]

So, a good day today. I've done the dovetailing to register the mould with the hard-shell (plaster/mod-roc) which will cover it, and very carefully prised the silicone from the pattern almost to the point of release at the top, before letting it close about the pattern again. Inspecting it inside out. This is to hopefully make it easier to separate the silicone from the base-board once the shell is applied, reducing the overall force needed to separate the plaster shell from the base-board in turn. This all being a new process to me it's a case of lots of research, talking to people who have done it before, and lots of thought before each stage.

Pleasingly I could see not evidence of either air-bubbles in the internal face of the silicone mould, nor any un-catalysed silicone remaining sticky rather than cured. This had been a bit of a worry, as after the 1st coat I'd noticed some slight "veins" of uncatalyzed silicone. I decided at that point to proceed and hope for the best, in the hope enough catalyst had mixed in the veins to nevertheless cure it, despite appearances. This proved to be the case, and with £150 quids-worth of silicone at stake, was a most welcome discovery. If the costs are high, bear in mind that to get this done commercially, I was looking at ten times the cost!

So, now that the silicone is cured and de-bonded from the varnished base-board, the next phase is a good covering of mod-roc (plaster infused bandage strip) before some thick layers of Plaster of Paris, and then a final binding with Mod-roc, which will give an overall thickness of the hard-shell of around 2 inches. In order to reduce the possibility of the plaster cracking in the exothermic reaction as it hardens, I'll probably need to do that in several hits, rather than all at once. The final stage will be to incorporate some sort of feet or runner's into the plaster, so that when inverted, the casting mould can sit level.

The moulding of the vac-forming pattern in alloy-powder infused epoxy resin can then begin. I'm aiming for an inch thick lay-up, backed with a generous amount of chopped glass-fibre, which should result in a polishable "metal" mould ideal for vac-forming and super smooth. The latter is critical, is the whole point of this is to end up with an optically clear cupola, so that footage taken from cameras within can show the outside world un-deformed by imperfections on the mould being rendered in the hot-pulled acrylic. It's "shooting for the moon" a bit, as this whole process from start to finish is new to me, although I've done carbon-fibre work in the past.

Wish me luck!

 

[Fidd88]

… and I've just uploaded the next film covering this:

Initial Plaster hard shell for mould

[Fidd88]

The last few days have been spent drying the plaster, using a fan-heater and a de-humidifier in concert, I appear to have been wildly optimistic in thinking it would harden completely in two days! So instead I took a trip to Alchemie, purveyors of epoxy resin with infused with alloy powder, and talking over the cast with their R and D chap, who was enormously helpful. The current intention is to demould the original pattern, wax it, then use it to displace the casting resin so that we end up with a cast "shell" of uniform thickness, circa 1 inch. This should reduce the chance of damage through too much heat being given off as it hardens, as well as reducing the quantity of resin needed. (it can self-damage if too great a thickness of resin is cured in one go)

On Monday I'm off to a local vac-forming company, who will do the actual pulls, to check that what I produce in the casting is well suited to the process. Then will come sourcing the very thin acrylic in 0.5mm and 1mm sheet. More than 1mm won't fit, but 0.5mm may be just too insubstantial, so a certain amount of trial and error is required.

In the next day or two, once the plaster is fully dry, I'll be putting another thick-coat of plaster on top of the plaster-bandage currently encapsulating the mould, before finishing it off with another layer of bandage. I'm hoping this layered approach will result in a good strong and thick support for the silicone component of the mould. Getting it all to separate from the MDF board is going to be good sport!

Overall it's been a fascinating set of new skills to learn, although I'd not recommend making such a large expensive mould as one's first foray into using silicone/mould-making! Still, if it all works, and I get the pair of turrets cupolas glazed with correctly curved and optically clear acrylic, I shall be well chuffed, and it's given me a lot of confidence of future mould-making on the Wellington itself. I can see that on a smaller-scale, a lot of these processes would be useful for making home-made specialist resin parts for your kits, especially if several of you banded together to share the costs of materials and to teach each other...

[Fidd88]

The saga of mould-making to vac-form the compound-curved windows for the turret cupolas continues:

Today I visited a local vac-forming company, who were brilliant in spending the time to help me understand the process and heat/forces involved. The upshot of these discussions is that I need to in corporate a honey-comb internal structure, even with the resin at circa 1 1/2" thick, as the cast as to withstand a "weight" of 1400 lbs over the 100 square inches or so of it's plan form. The easiest way to achieve this, and to economise on resin, will be to 3d print a series of hollow hexagonal plugs, linked at the top (ie above the level to which the resin is poured) and to withdraw these post cure, leaving hexagonal holes in the resin down to a distance of not less than  1 1/2" from the exterior face. I'll also radius these so that the load is spread to the walls of the hexagons, a bit like a vaulted ceiling. This male mould be simply be inserted once resin is in the female mould, and cramped into position displacing enough resin to bring the level to the top of the mould.

Since my last post, the mould and plaster shell were removed from the varnished base-board, and the interior of the silicone mould inspected fully, no bubbles, tears, stray hairs nor any other horrors. So far so good therefore! Film here: mould inverted and separated - no flaws!

[airscale]

OMG this is next level stuff

 

..can't be that far removed from making a real one!

 

Peter

[Fidd88]

Cheers, following advice from a local vac-forming company, the cast resin needs to be circa 2" thick. Consequently I've designed a 2nd male mould to be 3d printed, the function of which is to displace resin from the mould, reducing the amount required, but to be removed post-cure to leave tapered cylindrical holes with round bottoms in the resin cast. These should transfer crushing loads to the walls of resin between each cylinder, in the manner of a vaulted mediaeval roof, and so help resist the tendency of the mould to collapse under full vacuum, which generates 1400lbs of force over the whole mould!

Incorporated in the male mould are two features, loops designed to keep the male moulds separate from each other, which will be snipped off post-cure, and shaped deep slots In the top, to allow them to be spun in situ with a screwdriver to aid in demoulding. I've designed these myself, so hopefully they'll work!

View of the "top" of the male mould (during casting process) which is actually the underside of the casting when cured.

[Fidd88]

Having costed-out the male mould pictured above, I had to devise a simpler and cheaper method. This has now been done, and the resin etc ordered. Instead the ABS pattern used to make the mould will be partially submerged in the resin to displace it, giving a wall-thickness of circa 50mm for the eventual vac-forming tool. I'll need to make a wooden structure to hold it down, to offset any buoyancy it may have before the resin cures, and to position it correctly. Tomorrows job.

In the meantime, I  got sent a ton of material from a chap in Newcastle, including an .obj drawing of an FN4 four-gun turret. This is effectively drawings of parts that are 2d, so I'm using it as a basis to create a drawing of parts suitable for production, eventually as a kit, printable in any scale between 1:1 and 1:10. It's fun to be working on something other than an FN5! I'm applying everything I learned doing the FN5, so the basic drawing has no holes or threads, but hole centres are marked. This will enable the scale to be changed to the customer's needs without having to rescale all the holes and sheet material thicknesses to metric integers, or decimals thereof.

The pic below shews progress, but it's a little misleading, as most of it is the 2d wireframe, rather than the 3d parts. After an evening at it, the guns, gun-sight and sight-bar are done. Lots still to do! Eventually I hope to add the principle Boulton-Paul turrets, and the FN's 5a, 50, 25, 17, 20 and 120 before embarking on American turrets. 

[Fidd88]

This too will eventually be a 3d-printed model turret kit. Many many hours are now required of CAD, to translate the wireframe into .step files of individual components, and a great deal of work checking the geometry. The FN25 was not a success, although technically it was very advanced. It had two hydraulic rams in the turret, for elevation, and a hydraulic pinion type motor for traverse. Additionally it had two rams to retract and deploy the turret, The gunner stepped down into it once it was deployed, "being very careful not to kick the jettison lever as he does so" (!) - from the manual. Although it defended the lower hemisphere well in theory, in practice the limited apertures for windows made acquiring and tracking fast moving targets all but impossible, whilst the deployment slowed the aircraft down by 15 knots just as all speed was needed!

I believe an FN25 was supplied to the USA, along with other specimen British turrets, which kick-started US efforts at turret design. It was probably accompanied by a note indicating the issues it suffered from in service, which is likely why Sperry were able to make such a good fist of their ball-turret, having seen a "here's how not to do it" example to look at.

The FN25 was later adapted to carry, instead of guns and a gunner etc, a single 24 inch naval searchlight, and used in conjunction with ASV radar to surprise and attack surfaced U-boats recharging their batteries. It was so successful at this, that the Germans reversed their previous procedure, recharging in daylight when they could see enemy aircraft, and submerging at night to remain hidden when they couldn't. Although 27 U-boats were sunk using the "Leigh Light" as it was known, even it it hasn't sunk one, it was hugely influential, as U-boats seen diving before an attack, in daylight had their position and course plotted making them much more likely to be subsequently attacked and sunk by destroyers or other aircraft the following day. It also kept the U-boats submerged for longer, reducing their range and speed.

Out of curiosity, if I released these turrets at circa 1:4 scale, as a pneumatically operable standalone model, probably made with resin parts, would there be any interest here making them, and if so, what might a comparably-sized resin model cost to buy? (roughly the size of a soccer-ball)

[Fidd88]

So, I put my rather questionable wood-working skills to use today making a box to manage the top-heavy plaster/silicone mould which will shortly be used to pour the resin tool for vac-forming the cupola windows. Blimey what a performance! Hopefully this will side-step the possible eventuality of the mould falling over and depositing 150 quids-worth of curing resin all over my workbench/tools/my feet/the floor. This would not have brightened my day. So, after a good clear-up tomorrow to de-dust the workshop as far as possible, the resin-pour should happen soon. 

The resin needs a complicated post cure heat treatment to strengthen it. I have discovered it will just fit in my kitchen oven. The next task is to figure out thermometer which can be inside the oven, but readable outside, ideally working around whatever damnfool interlock is fitted so one can't run the oven with the door open...

latest installment: