SimonBrown

Tool appeal: Does anyone have a 5/16 BSW or 3/8 BSF crows foot spanner lurking in their toolbox they would be willing to sell? 1/4" or 3/8" drive would be perfect. The cylinder bolts are difficult to access and whilst a ring spanner is about to go under the grinder, a crows foot spanner would be a great help. Even the internet agrees they are a rare thing - couldn't find one - but folks here might just have one not required and lurking in the depths of the toolbox?

It was time to stop fussing around and get the manifold off. Plenty of heat, aided by youngest daughter directing the blow torch onto the coupling, and a long extension tube on the C spanner did the trick. Finally shifted them and (crucially) with no damage. Wrestling one of the sections off the motor was fiddly, and it going to be fun trying to get that one back on. The complexity may well have been the killer of this idea. There is plenty of power in a very compact space, but the manufacturing and assembly plus any in-service work would have been expensive. Glad I took the manifold off. Found my first bit of rusty crud in one of the inlet ports: With the manifold out of the way the rest of the fuel pipes came off. Some route to/from what I thought was the fuel distributor, but I am wondering if they are not spill/return pipes? More investigation needed. A bonus was finding inspection stamps on each of the cylinder castings, just next to the integral fuel injectors: The only other time I have seen stamps akin to this are on aircraft components. Presumably stamped at the Whitehead works in Weymouth, they need following up at some point.

With a single original track segment it could be scanned using photogrammetry - just like the torpedo motor I'm restoring - with the advantage it would be easy to make it a closed & therefore solid model. It would need work post-scan to add material to any worn sections, add material to the track pin bore and scale it to take into account the shrink factor when cooled, but the beauty of a 3D model is much can be tweaked and changed at very little cost and verified with a 3D print before committing. More than happy to help on this one, if photogrammetry is of use?

It took longer to assemble the borrowed hoist than it did to get the engine onto its stand: The inlet manifold threaded connectors are proving to be a PITA. Three of them have come undone no bother, but the rest are tight AF and are refusing to budge even with an extension bar on the C spanner and applying loads of heat. Its as if they were welded up but there's not sign of that. I'm heating them with a blowtorch and allowing oil to run into the threads as they cool, but if anyone has any other ideas then feel free to chip in: I couldn't resist a peek into the engine internals so I pulled the auxiliary drive cover off...Wow was the first reaction: Everything is numbered to its relevant pot, which will make reassembly easier. Plus, it looks like we have an engine number of sorts: 113 is stamped on every component I can see so far. I had seen this stamped on one of the poppet valve covers and wondered its significance, but with everything carrying the same number + cylinder number its a logical conclusion: Degrees stamped on the cam presumably relate to valve timing, but will verify at some point. After drooling over the engineering I slipped the cover back into place and carried on heating and soaking the inlet manifold glands.

A few hours spare over the last couple of weeks has got the adaptor sorted for the engine stand. Closer inspection of the crankcase housing reveals its an aluminium machined block, not a casting. It will be interesting to get the aux end cover off and peek inside. I got some steel blanks plasma cut. Ordered online to the exact size I wanted, including the flange: Holes needed marking and drilling but the steel blanks were perfect and burr free. Quick loose fit of the flange confirmed the marking out was close enough. And the main mounting plate lined up too. After tack welding the spacers in-situ I took it off the engine to put a bit more weld around each. With my welding I was not planning to rely on the strength of the fillets, but by weld No 28 of 32 quality was improving: And final fitting, ready for a lift onto the stand: Next few days should see the engine off the floor and onto the engine stand. Then its back to stripping the motor down.

At the point I was pulling this bit apart I wasn't really sure a) what it was and b) how it came apart. With no manual it became a bit of trial and error until I realised the distributor barrel was machined to very fine tolerances and it was 50~70 year old gummy fuel residue putting up resistance... but until that point the last thing I wanted to do was force something and inflict damage what may be the last of its kind. In the end, it took a gentle but persistent push with some penetrating fluid down the bore and ports and it slid out with nothing more than finger pressure.

A little progress this weekend. The rusted split pin was drilled out and the last set of fuel timing gears were freed from their bronze casting : Two 'C' spanners were welded together to fit the exhaust coupling slots better - but I'm not sharing photos of that. Its been 30 years since I last used an arc welder and lets just say it shows. The welding is good enough to hold them together and work as a pair, but not quite display material. Friday the engine stand I ordered turned up. Whoever designed it had never envisaged it being used to support an 8-cylinder radial 2-stroke diesel engine, so some lateral thinking/cutting/welding will be needed to hang the engine off it and not cause damage to the aluminium crankcase in the process. But if I'm being honest, its solving problems like this that makes it all the more rewarding.

With an hour spare, more progress. The bronze lump on the aux drive end is almost certainly for fuel distributor/injection timing. With a bit of penetrating oil, lateral thinking and gentle pressure the large cylindrical barrel slid out of the housing. The barrel has a series of ports, each of which have a sliding block (the copper-looking rectangular bits in the image above) that marry up to an array of ports that have been drilled into the bronze liner thats been pressed into the casting. You can just see the bronze liner above. From what I can tell - and I need to stick some fluid through it to prove it - the four ports at the bottom (you can see three above) are feeding pairs of cylinders. The exact timing is unknown. The rest of the gear assembly came apart a treat, with just one rusted split pin refusing to budge. That will need careful drilling next.

The hydraulic platform warranty repair arrived this morning. Lots of excitement, with all hopes dashed when the load bed refused to lower when anything of any mass was placed on it. Back to square one with a refund now heading my way. Disappointment did not last when the postman arrived with a package containing a set of Britool BSW sockets. Shiny, quality tools are always a joy to work with. The 'C' spanner arrived too, so it was time to start work on stripping the engine. The inlet manifold is made up of several sections and the 'C' spanner managed to undo the first 3 couplings no bother. The joins are sealed with some very fancy copper washers. The rest of the unions are stuck fast and refused to budge, even with a little percussive maintenance...so they are soaking in penetrating oil now while I acquire a decent blowtorch and get a bit of heat into them The sockets went to work on the aux drive end of things - the bronze yet-to-be-identified parts at the rear. What I thought was an oil pump may well not prove to be so after stripping. Its a geared, two speed affair with a set of unidirectional dog clutches. Its purpose remains unknown...however, sitting under this body was a gear pump. The galleries leading from this pump lead deeper into the engine body, so at this stage I'm guessing its for engine lube. I'm not sure there is a sump and it reminds me of a JAP speedway engine fitted with a pilgrim pump, so called as the oil goes just one way so my dad told me. The design of the gear pump does now tell me rotational direction of the engine, if nothing else at this stage. Next up, the bronze end cover/oil pump housing came off to reveal the outer end of the crank main bearing. From what I can see its looking near-brand new inside, which means we are not dealing with a seawater flooded pile of internal rust. Overall, the engineering is pretty impressive for a motor designed to function just a few times before its terminal destruction.

This policy shall be the guide I think. Thankfully this engine has not been in salt water for a long time - if ever. Its only light surface corrosion so far. I lifted a cannon ball from a wreck site we found on Chesil Beach after winter storms uncovered the pile of cannon and shot. That was about 5 years ago and as far as I know its still in a fresh water bath, encouraging the 300 years of salts out of the iron and back into solution. Had the engine been raised from the seabed I think it would be in a far worse shape. That sounds just the ticket. I will have a hunt around - cheers!

Thanks for the measurement. The manifold inlet on the 8-cylinder is approx 1" dia, so its a useful comparison and it triggered a read of the archive docs. From the records I found performance specs from two 8-cylinder engines, with swept volumes of 406 cu in (6.65l) and 460 cu in (7.53l) respectively. The former was listed as built, the latter listed 'design delayed'. When Britain entered the war effort shifted from R&D to production, such that by 1943 we had more torpedoes than we needed, but it was acknowledged any R&D would not be in service until the war ended. Air consumption for the 8-cylinder that had run is listed at 269lbs of air - which converts to approx 94,374l of air for a 60 knot run for 148 seconds/5000 yards! . By comparison the standard 4-cylinder Mk. VIII & IX 21 inch engines consumed 141lbs (49,467l) of air for a 45 knot run for 197 seconds/5000 yards. A typical 15l dive cylinder at 232bar will hold 3480l of air...so if my maths is right we won't be running the motor for too long. Which considering its age and rarity is perhaps no bad thing? On the other hand, divers like myself use enriched air - breathing gas mix with anything between 21%~100% O2 - for reducing risks with decompression sickness...and the Japanese did indeed put enriched air torpedoes into service...given the risks dealing with high pressure O2 this idea will need to go through the "Can I? Should I?" filter more than a few times I think. I digress. I really am in two minds with what to do? Preserve the patina or restore to look new? Only get one bite at this cherry and once done there is no going back. Preserve is now under consideration. Could you share a link to such products? The main body is aluminium.. There is a brass component on the aux end, its purpose currently unknown but is (I think) either a lubrication pump or an incomplete fuel pump. It does have a nice patina...I have created some highly detailed images (called orthophotos - 1mm per pixel is a typical scale, great for recording stuff under water) that act as a permanent record of the motor before I touch it. The aux view of the engine can be seen near the bottom of this link: Orthophotos of torpedo engine. So...keep the patina? Or return to the as-manufactured look?

Well, the good news is the garage is well stocked with dive kit, and 250bar cylinders are in abundance.

One step forward...one back. The hydraulic table platform arrived today. Its going to save my back no end and with space at a premium in the garage I can move it around...all fine...got the engine on the load bed and raised it...only to find the damm thing would then not lower. No amount of fiddling with the lever resolved it. After a bit of head scratching I managed to get the motor off the load bed, and sure enough it lowered when the lever was pulled. A quick test, by sitting on it and trying to lower myself, repeated the problem. So its back to the supplier I fear. Starting to plan the first steps one job will be getting the injector pipes rust-free. I am loathed to use wire wool or other abrasive methods, so has anyone suggestions on how to clean the rust off without abrading the whole pipe? Rusty pipe and painted example:

A trip to the National Archives on Friday was reasonably productive. The 21 inch 8-Cylinder technical report was viewed and it actually contained a couple of photographs of the test unit. Whilst similar in design it was apparent (from the cylinder head bolt pattern) the engines were not identical. There were other reports that mentioned an 8-cylinder engine and have a bit of reading to do - all interesting stuff. But one thing jumped out of the tech report. The engine was designed to produce (and tested to) a rather impressive 800hp. I don't see any reason why my example would not be equally capable?

Now that did bring back a few memories. Spent a few weeks in the summer covering Crazy Horse as it was used to conduct trials on how much delay an operator could cope with on the TV signal. Thanks for the reminder.

American record cards are indeed month/day/year - that possibility has been taken into account.

RARDE Chertsey developed a radio controlled Chieftain back in the late 80s/early 90s. Known as Crazy Horse, the Chieftain was controlled from a Spartan APC. It did need an operator, and was remote control rather than autonomous. If memory serves me well it did feature in a TV program at the time...but cannot remember the details. Apparently its in storage now at the Tank Museum down in Bovington.

Well this has not been something we previously considered. Of the two sets of records - RN and USAF - the latter has (anecdotally) been acknowledged to be occasionally lacking detail and accuracy. We have taken the RN record at face value. I was in the National Archives again today and double-checked the entry, just to make sure (have a horrible habit of transposing numbers...) but it was the 7th of May. That aside, its worth looking again at the P-47D records and see if we have anything matching - worth following up so thanks for the lateral thinking.

No, and no - serialisation of major components and their attachment to airframes hadn't gone to the later levels of detail. Knowing the engine number won't help, only the manufacturers or USAF serial ID. We did spend an entire dive having a hunt for the engine data plate but no joy. The radial engine casings have fizzed away in the salt water and the conclusion is said plate has gone the same way.

It kind of is, and isn't at the same time. We would love to ID the aircraft but as its taken somewhere between 15 & 20 years to get to where we are now, patience really is a virtue. The sea never gives up its secrets lightly. We thought we had a serial number match recently and requested the record card from the USAF Historic Branch, but when it arrived it was apparent that the ID was for an aircraft that was written off thanks to non-battle damage and was salvaged. Clearly ours wasn't salvaged, plus the recorded date of loss didn't match. Thanks for the link. It might be worth dropping them an email and ask if they could take a look. Apart from the data plate only the main wing spars were stamped with the serial number, but anecdotally we hear the control linkages were informally marked once they had been set/adjusted on the build line. It would be handy to confirm that.

Like most things related to cost, its all a question of what you really need as a result. The cost is typically in the detail needed. First up, the camera. I have used my iPhone to scan, and have my full frame DSLR. The iPhone is almost a perfect camera, but only by the measure of its in your pocket when you need it. When it comes to pixel density - the thing that drives accuracy and level of detail in a 3D model - the DSLR wins hands down and is used every time I'm underwater (most of my work is underwater stuff) or when I want a detailed record like the motor here. The gap between the cost of a DSLR and an iPhone seems to be closing, but even now the DSLR costs more. It really depends on detail needed. More detail = more cost. Then the software to process the models comes in two forms; Basic and Pro. The basic one will render a lifelike 3D object but it will lack accurate scale. If you wanted to scan a section of the Atlantic Wall then the Pro version will not only allow you to scale it for accuracy, but to also embed geographic data to then reference the model in the real world. The Pro version can also create what is called an ortho photo and a Digital Elevation Model. The ortho photo is like a site plan, with a scale of 1mm per pixel (typically) and the DEM shows height in colours - this link will show more P-47D Crash Site. There is approx 10x cost difference between the two versions of software and for scaling alone I have the Pro. If you want to print or model a replacement part from the 3D scan then scaling will be needed. Then there is the hardware to process the models on. More detail = more time, or a faster computer. The processing of 3D data really benefits from a graphics card, which are normally as much as a decent laptop in their own right. Lots of RAM is needed too, plus the fastest processor you can buy will typically help. Or you can use older hardware and wait a long time, or work with much reduced detail. One job I did for the University of Nottingham was to scan the wreck of the SS Thistlegorm in the Red Sea. 24,000 images were needed to scan the wreck and its military cargo of trucks and ammunition, and even with some very costly cloud processing it still took 65 days of computer time to finish the job. The hardware I use to process 3D data is not cheap: Around £4k for a capable PC that does 99% of what I need in a timeframe I can live with. I rent time in the cloud when I really need more capacity for larger models. By now you may well have guessed its a trade/job for me. Work is not regular, but there are very few people doing this kind of work underwater and I have delivered two jobs for National Geographic TV Channel - one on Port Royal in Jamaica and one in Greece looking at the seabed and wondering if it was man-made or not. I have access to a drone too, but rarely use this for 3D work as there are others already doing it. My real love and passion is history thats underwater, so being able to scan and bring back models of what I see and share it is very rewarding. Last October I dived in Scapa Flow and paid a sonar target a visit...thus becoming the first in 100 years to see a battleship anchor from the German High Seas Fleet. With photogrammetry (the technique) means I can let everyone else know what it looks like - scroll down on this page if interested. So is this 3D stuff in in reach of everyone? Short answer is 'yes' if you have a PC, camera and the software (Agisoft is the company and they have a 30 day trial download). Its quite easy to get a basic model done in a short time, but the detail and accuracy will always take a little longer to master. I do think there is an application to recreate or copy the hard-to-find/impossible-to-replace things that all sorts of vehicles need. 3D scanning is just one step on a path than leads via 3D printing to lost wax casting in low volumes, or machined replacements.

Well the first step was to scan it into a 3D model. Should be a great reference and reminder when it comes to reassembly. The poppet valve covers have an 8-splined socket machined into them. Looks like someone went to a lot of trouble to make that. So far, that looks like the only special tool needed, but I have a sneaking feeling this will change as it comes apart. The poppet valve covers are also numbered 1 to 8. So far, no other numbers found. Link to the model here - its quite complex and will take a while to load.

Good spot Richard, it was already on my reading list at the NA for next year - but the link much appreciated as I could have equally missed it. The friend of mine still has a few much older mystery torpedo parts - if you turn up anything in the Collingwood collection relating to Whitehead torpedoes then do please let us know. Another one I was going to follow up on, but it looks like they are suspending access to archive material while they move their collection. Thankfully there will be the chance for a face-to-face "Ask and Archivist" day so will keep an eye on those. Me too! To think such a sound has not been heard for a very long time...8-cylinder 2-stroke...open pipe...

The potential for creativity unleashed is almost without limits, you could say.

Its the cooling that is occupying my thoughts and with no apparent resolution at present. Seawater was indeed used, flooding the engine compartment negated the need for any other cooling.