Best tank engines of WWII

[WARSPITER]

At 990 lbs. couldn't you do an R-1820 and realize a 700-900 hp powerplant to move your tank for roughly the same engine weight?

Why not ? It was a proven product and maybe a tight squeeze for some tanks but it would also give scope for higher output
if needed. It worked ok with the Merlin.

 

[Shortround6]

The R-1820 wasn't quite that light and it was bulky, which is more important.
M3s and M4s grew in length but fitting in a 55in dia engine instead of the 45/46in diameter Whirlwind engine used in the M3 & M4 was a bigger problem.
2nd problem, which was huge, was trying to replace a 975 cu in engine with 1820 cu in engine with roughly double the toque. This is like swapping out a 2 liter car engine for a 4 liter engine and trying to use the original transmission. It will work.................for a little while.
Change in clutches and transmission is going to be much heavier than the difference in engine weight.

 

[Elan Vital]

R-1820 was fitted in tanks as the G200 for the M6 heavy tank. A dieselised D200/RD-1820 was fitted on the M4A6, and regular G200 was contemplated for a M4 mounting.

 

[GrauGeist]

R-1820 was fitted in tanks as the G200 for the M6 heavy tank. A dieselised D200/RD-1820 was fitted on the M4A6, and regular G200 was contemplated for a M4 mounting.

The M6/M6A1 heavy tank project reached a dead end and the M4A6 development was scratched because of compatability issues.

 

[Reluctant Poster]

An even more direct successor to the Jumo 205, which had been in service for decades, was the Kharkiv-built 5TD(F)/6TD mounted on the T-64.

Dare I mention the 'legendary" Leyland L60? It was a contemporary of the 5TD.

 

[ThomasP]

In all seriousness , by the time I was working for Teledyne/Chrysler on the M1 Abrams Alternate Diesel Powerplant Project, the L60 was a pretty good engine. The design teams had looked at the engine, not as a powerplant for the M1 - its output (750 BHP) was nowhere near the power output they were looking for (1500 BHP) - but as a possible design type for development. The L60 was already a multi-fuel engine, and the opposed-piston layout promised a small package, and by 1979-85 it was already a reliable engine. No idea what the final conclusion was as to whether the design type should have been pursued or not, due to the decision to go with a US made engine.

 

[Reluctant Poster]

This sub-thread needs to fade away but I wanted to "go on the record" to show in a simple diagram the point I was trying to get across. (Please excuse the "chicken-scratch" handwriting of mine in the labels on the diagram, but I had hand surgery five days ago and still have bandages on my writing hand.)

This diagram has a target (T) that is a Victory ship traveling at 18kts. The submarine (S) can make 17kts at best. (Typical for a VIIC doing long runs.). The sub is running on the surface. This one-knot advantage has the following effect: If the sub spots the target (moving to the right) at a point on the perpendicular of the target's track, or further to the right on that track, the sub will never be able to approach the target to get into a firing position, unless the target passes very close to the sub. This is the simple geometry of the problem. Unless the target's track passes within a short distance to the sub's present position, it cannot get into a reasonable firing position. This is because the sub has to track and approach the target on a hypotenuse of the right triangle (the longest "side") formed by the target's track, the perpendicular line to the target's track, and the sub's direct track (hypotenuse). Since the sub's track will always be longer that the target's track, the sub will never be able to "catch up" to the point he can attack at a reasonable range and angle. If the target has an even larger advantage (e.g. warship running at 20kts) the hopelessness of the situation is apparent. (This is why reminiscent books of WWII submariners and ships logs, etc. are replete with situations where the sub "could not approach target and get into firing position.") Running down the hypotenuse means that the sub's velocity vector has a component in the direction of the target track as well as a "forward" component to get ahead of (or equal to) the target's position. The target's vector, on the other hand, is a straight line down his chosen track.

There are 3 scenarios in the diagram based on the perpendicular distance to the target when spotted: 12nm, 8nm, 4nm. (The horizon for the top of cargo ship from a sub's coning tower is about 12-13nm in daylight, by my calculation, although this varies greatly depending on sea state, mist, fog, etc.) The position of the target is the tic-mark on the target track with the time and travel distance along his own track. The semi-circle is a 3000-yard "firing envelope" for the target (about 1.5nm). I chose this number simply because the examples in the US submarine torpedo firing manual rarely list any distance beyond this. The little "dot" in the semi-circle is half this distance perpendicular to the target, or about 1500yds. This is where the sub captain will "aim" his vector or, rather, where I draw his vector aiming here. This would be his "ideal" firing position, if the sub can reach this. The sub's position at this time is the tic mark at the end of its vector line.

Note that the longer the perpendicular distance to the target's track the greater the sub will eventually lag behind the firing range semi-circle. Conversely, the shorter the perpendicular distance to the target's track when spotted the more likely the sub will get into a firing position within the semi-circle. In the 4nm case, the sub can just get to the edge of the semi-circle (3000yds from target). At shorter distances than 4nm it will have an even better chance to "shoot." (This is because the distance to the track is smaller and, thus, the hypotenuse's length is closer to the target track's length.)

The diagram was made with meticulous attention to scale accuracy. (Note the smudges where I was off by 1 or 2mm and had to re-draw.) Looking at the big picture, imagine a submarine in the center of a big circle whose radius encompasses the sub's ability to spot a target at any direction. The actual angle that the target is approaching is immaterial, as the diagram is circularly symmetric and you just have to draw a line from the sub perpendicular to the target track and rotate the diagram sheet and you have the equivalent problem. The above arguments mean this: If the target has a speed advantage over the submarine, then the right two quadrants (right semi-circle of the big circle) are off limits as to the possibility of approaching the target to get into a good firing position. The greater the target speed advantage the closer its track must be to the sub position in order to get even a chance at a torpedo shot. The German's understood this fact well and is why they worried that, eventually, the majority of convoys might be all 17-18 knot Victory ships which would significantly reduce their success rate at tracking and attacking a convoy due to simple geometry. In this case their subs would have to spot the target in the "left" half semi-circle of the maximum spotting whole circle. Due to the vastness of the ocean and the randomness of finding convoys, this could not always be guaranteed and would probably not be the case in about 50% of the time, due vagaries of navigation and the state of the sea. They would always be unable to get in front of the target (convoy) if first spotted in the right semi-circle. In some cases this would apply as well to targets spotted well into the left spotting semi-circle, if the speed of the targets are much more than 1 knot greater than the sub's maximum speed and their track is perpendicularly far enough away from the sub.

View attachment 874814

Lets get the 17 Knot Victory ship out of the way . That was not a deliberate design choice, it was an accident of history. The US had a significant mismatch in their ship building program. They had substantially over produced the 8,500 HP power plants for the C-3 cargo ship program. This was a very advanced powered plant that was not really suited for what was intended to be a mass produced ship. However the units were going into storage and it was decided that rather than waste them to install them in the Victory ships. That is why early production Victories had the more powerful and complex power plant. After the surplus units were used up, later production switched to the intended powerplant.

There is a myth that the victory ship was given 15 knots speed to escape the U-Boats. In reality the US had determined that the minimum speed for cargo ships to be at least 15 knots going back to the original design of the C-2 cargo vessel back in 1937. The US never liked the Liberty ship which they considered it to be a 5 year ship. They really wanted to build more of their excellent C-2s but production bottle necks in turbines and particularly reduction gears meant that had no real choice but to build Liberties. As the turbine and gear situation eased the victory ship was to be built in the emergency ship yards to supplement (not replace) the C-2

The C-2 serves as an excellent stand-in for as to how the Victory ships would have performed pre Black May. 7 were lost out of 44 in cargo service in 1942 and 1943.. Note that note that not all of these served in the Atlantic. I found 2 that were definitely in Pacific.

U-boat net gives details on the individual sinkings. From: Ship types hit by U-boats during WWII - Fighting the U-boats - uboat.net

At 16.24 hours on 9 July 1942 the unescorted Santa Rita (Master Henry R. Stephenson) was hit by one torpedo from U-172, while steaming on a nonevasive course at 16 knots about 700 miles northeast of Puerto Rico. Lookouts had spotted the wake but it was too late and the torpedo struck on the port side between #3 hatch and the engine room

At 15.43 hours on 16 July 1942, U-161 attacked convoy AS-4 about 500 miles north of St. Thomas, Virgin Islands and observed two hits on a first ship after 2 minutes 32 seconds and heard a third detonation after 3 minutes 35 seconds. Achilles reported one ship sunk and another possible damaged. In fact, only the Fairport was sunk by two torpedoes

At 20.25 hours on 9 January 1943 the unescorted Louise Lykes (Master Edwin John Madden) was hit by two of four torpedoes fired by U-384 from a distance of 1800 meters, while proceeding on a zigzag course at 15 knots.

At 02.51 hours on 24 April 1943 the unescorted Santa Catalina (Master Olaf Berg) was hit on the starboard side by two stern torpedoes from U-129, while steaming on a zigzag course at 16.5 knots about 370 miles southeast of Cape Hatteras, North Carolina.

At 06.56 hours on 12 July 1943 the unescorted and zigzagging African Star (Master John George Waller) was hit by a torpedo from U-172, which struck the port side at the #4 hold.

At 09.38 hours on 17 March 1944, U-371 fired a Gnat at convoy SNF-17 about 30 miles north-northeast of Bougie and observed a hit on a ship, which settled by the stern after the hit. At 09.42 hours, a spread of three torpedoes were fired and two hits were heard. After another Gnat at 09.48 hours, a further detonation on another ship was heard. The first torpedo struck the Maiden Creek in station #52 and the second the Dempo. The first ship was hit again at 13.50 hours by a coup de grâce.

Note that were the ship speed is given, it is above 15 knots

I did the same exercise to the for the fast tankers that the US started building before the war including the famous T-2s and counted at least 15 torpedoed by U-boats. More examples from U-boat net:

At 06.36 hours on 23 Nov 1942 the unescorted Caddo (Master Paul B. Muller) was torpedoed by U-518, while proceeding on a zigzagging course, changing every six to nine minutes. The torpedo struck the port side at the pump room, just forward of the after bulkhead.

At 20.00 hours on 10 June 1943 the unescorted Esso Gettysburg (Master Peder A. Johnson, lost) was hit by two torpedoes from U-66 about 100 miles southeast of Savannah, Georgia, shortly after she received a U-boat warning, steaming on a zigzag course at 15.5 knots.

At 00.08 hours on 2 July 1943 the unescorted Bloody Marsh (Master Albert Harrison Barnes) was on her maiden voyage, when the ship´s torpedo indicator sounded after detecting the approach of a torpedo from U-66. The master ordered a course change to hard left, but 30 seconds later the torpedo struck the port side at the engine room, destroying the room completely, flooding the compartment and killing one officer and two men on watch below.

At 09.21 hours on 22 July 1943 the unescorted Cherry Valley (Master John H. Rose) was hit by two torpedoes from U-66, while steering a zigzagging course at 15.5 knots.

At 07.47 hours on 3 Dec 1943 the unescorted Touchet (Master Jesse Field Bird) was hit on the port side by a torpedo from U-193 in the Gulf of Mexico while steaming at 16.5 knots.

At 10.21 hours on 16 Dec 1943 the unescorted McDowell (Master Henry David Barrow) was hit by a Gnat from U-516 about 30 miles north of Aruba. The torpedo struck the port side at the screw, destroying the propeller and shaft and disabling the main and auxiliary engines and the steering gear

At 07.25 hours on 26 Dec 1943 the unescorted Chapultepec (Master Allen L. Remick Jr.) was spotted in hazy weather by U-530 about 90 miles northeast of Cristobal. The torpedo fired was noticed by the tanker´s acoustic torpedo detector but seconds later struck just abaft the stem

The Kittanning (Master Raymond J.S. Chambers) had left port at 13.30 hours, but soon thereafter the third assistant engineer fell and seriously injured himself and the master decided to return to Cristobal. At 16.31 hours, just after the ship changed course back at 14.5 knots, a torpedo struck on the starboard side at the #6 tank.

At 20.02 hours local time on 2 Nov 1944 the unescorted Fort Lee (Master Ottar Marius Andersen) was hit by one torpedo from U-181, while she proceeded on a nonevasive course at 15.5 knots

U-Boat net also has a list of the largest ships sunk by U- Boats. Many of which were high speed vessels:

Largest ships hit - Ships hit by German U-boats of the Kriegsmarine - uboat.net

The U-boat War in World War Two (Kriegsmarine, 1939-1945) and World War One (Kaiserliche Marine, 1914-1918) and the Allied efforts to counter the threat. This section includes over 21.000 Allied Warships and over 11.000 Allied Commanders of WWII, from the US Navy, Royal Navy, Royal Canadian...

uboat.net

While the UK didn't have any fast tankers they did have a lot of fast cargo liners. I started looking through

Cargo Liners: An Illustrated History

by Ambrose Greenway (Author)

https://www.amazon.com/s?k=9781848321298&i=stripbooks&linkCode

I lost count but suffice to say a lot of fast British ships were lost to U-Boats.

USN submarines were faster than U-Boats but the Japaneses tankers were faster than the Ametcna ones. It would be intersing to see how many of the 19.5 knot tankers were sunk by subs.

https://www.combinedfleet.com/ItsukushimaM_c.htm.

As you can see high speed was no guarantee of safety. In fact after Black May very few ships sailing in convoy were sunk. The US realized convoying was essential and started running high speed tanker convoys (14.5 knot minimum) from the Caribbean to the Mediterranean and the UK as well as fast troopship convoys (15 Knots minimum) to the UK and Med. Very few ships were allowed to sail alone across the Atlantic by the time the Victory ships appeaerd.

 

[Elan Vital]

FWIW the French ARL tank design bureau released a paper on dedicated tank engines in 1939 that gives an idea of how things were theorized:

"On road, it is preferable of increasing average speed through increased engine rpm.

Experience shows that in this case the power needed for continuous operation doesn't exceed 70% of the maximum power the engine can output at this rpm.

On varied terrain, it is rarely necessary to use the maximum engine rpm but it must provide continuously a high ouput that can be estimated at about 90% of its maximum power with occasional torque spikes such that this engine runs at said maximum power.

If we define Wmax at nominal rpm the maximum output that could be held for 1 hour at the nominal regime;

Wnominal the maximum the output that may be held continuously at the nominal regime.

Wroad the output that may be held continuously at 115% of the nominal regime it will be necessary to have the following relationships:

a) between outputs: Wnominal=0.9*Wmax and Wroad=0.7*Wmax

b) between regimes: Nmax=1.15*Nnominal

Current engine technology allows these relationships.

The ARL then created a table giving engine power intervals for a given weight class of vehicles, defined to provide two advantages:

a) minimum specific consumption for the required performance

b) allows the manufacturers to use elements of existing engines for calculations

It was evaluated that given the required durability of the engines, technical possibilities and the potential to mount the new engines on existing vehicles, the required nominal rpms would be 2300 to 2500 rpm for 340 to 700 hp engines, and over 2500 rpm for 50 to 160 hp engines.

The table was as follows, with each column corresponding to maximum output at the nominal regime, preferred layout ("en ligne" meaning inline/straight engine), bore, stroke (which was fixed as 1.3 times the bore), length, width and height all in mm:

The 50hp I4 engine was suitable for the likes of the Renault UE infantry load carrier, the 120 to 150 hp I6 was suitable for light infantry tanks like the Hotchkiss H39, the 340 to 400hp V12 was suitable for the B1 Bis/Ter, and the 550 to 700 hp V12 (or flat 12s) would be suitable either in pairs for fortification assault tanks, or alone for heavy tanks.

The next table defined for each vehicle class the nominal output, maximum output at nominal regime, maximum output at maximum regime and road power, with the corresponding durations for bench trials in hours. Additionally, the engines were to be tested unloaded at 110% of the maximum regime for 10 minutes (I believe this is to test resistance to the occasional overspeeding when going down a slope).

These engines, beyond being optimized for military vehicles and giving greater power and durability/reliability than the existing engines which came from automobile or aircraft industries, also were to offer lower specific fuel consumption (240 g/hp/hr instead of 270) and much smaller dimensions relative to their displacement.

The 340 hp-class engine ended up being intended only for the ARL V1 75mm SPG (weighing around 25.6 tonnes on paper with such an engine).
The two entries for this class were:

- a Talbot V400 carburetted V12 with a displacement of 17.65L (120mm bore and 130mm stroke) and a CR of 6. At 2400 rpm it gave a continuous 300 hp and maximum 340 hp, while maximum torque was 93.5 mkg at 1200 rpm and torque at 2800 rpm was 86 mkg, all on 65 octane petrol.

The engine could be also be set to give 325 hp at 2400 rpm and 380 hp at 2800 rpm, with specific fuel consumption to remain under 270 g/hp/hr. There were two Zenith 60 ZA2 carbs, the 2 water pumps gave a flow rate of 140 L/min at 2800 rpm engine speed and 58 L/min at 1500 rpm. The cooling system was to ensure a temperature increase of less than 15° C between the engine inlet and outlet.

- a Brandt fuel-injected V12 with a bore of 110mm and stroke of 125mm. It didn't have the higher setting of the Talbot engine, but the fuel-injection allowed a reduced specific fuel consumption of 220 g/hp/hr and a reduced displacement of 14.26 liters. Dimensions were 1160 x 700 x 850mm L*W*H.

Renault designed an entire family of engines with a straight-4, straight-six and V12 around the same 130mm bore x 130mm stroke cylinder, but these skewed towards the higher end of ARL's power categories.
The V12 gave 420 hp at 2800 rpm (intended for the production B1 Ter with new transmission elements as a result, or the future battle tank) and 360 hp at 2400 rpm. Its dimensions were 1200 x 760 x 850 mm.
The straight-6 meant for the Renault DAC 1 light infantry tank (a 16-19-tonne design) delivered 180 hp at 2400 rpm and in theory 210 hp at 2800 rpm (200 hp without any associated rpm was cited in DAC 1 documentation).
The straight-4 delivered 120 hp at 2400 rpm and in theory 140 hp at 2800 rpm, but isn't associated with any specific vehicle.

Renault also designed a 155 x 165mm straigh-six for the B1 Ter and Renault G1R which gave 350 hp at 2500 rpm.

They were functionally middle-of-the-road in terms of power density and output, and somewhat similar/slightly more efficient than the Hispano-Suiza V12s I discussed previously. In comparison, the 325-350 hp Bedford engine in the Churchill (21L) was 1320 x 1220 x 750mm, the Matilda II's twin engine installation (diesel and inefficient twin engine setup) was 1450 x 1470 x 950mm, Valentine's (diesel) was 1200 x 820 x 900mm. The downrated 600 hp HL 230 P30 was 1115 x 892 x 900 mm. The Meteor was 1470 x 983 x 970 mm . Both the Maybach and Meteor gave a better output relative to displacement than the French engines.

However, this comparison says more about the relative inefficiency of early-war engines adapted from civilian types than the efficiency of the dedicated French designs. The Matilda II had relatively weak diesel engines mounted in pairs in a bulky setup, while the Bedford was a flat 12 and was not a high-output design.

Compared to the legacy French engines, the B1 Bis' Renault straight-six 140x180mm delivered 280hp with the intake/exhaust losses removed and 300 hp with ARL modifications and new pistons for a 16.6L engine, so the new Renault 155 x 165 straight-six increased specific output from 18 hp/L to about 19.8 hp/L on 65 instead of 72 octane fuel and with greater durability intended. The Talbot V12 is similar but with the V12 form factor and option for greater output. The Renault V12 massively increased available power although the specific output itself increased to just 20.3 hp/L, and was "more compact than the engine it was to replace".

The bored-out V8 engine in the SOMUA S40 delivered 220-230 hp out of 13.75 L (16.7 hp/L), so it was far surpassed by any of the new engines, especially the Brandt V12 which had a barely greater displacement of 14.2 liters but a 50%-greater output. The straight-six 130x 130mm Renault engine for the DAC 1 delivered almost the same output as the SOMUA on 10.35 liters.

 

[Shortround6]

Thank you.
what seems to be missing from these French calculations is the transmission set-ups.
You can get away with a lower powered engine IF you have more gears in the transmission. But that means a more expensive, larger, heavier transmission. You can also go too far in this direction as the Germans did. You also need to make sure such a transmission is easy to shift (at least somewhat) and that the drivers are trained to know when to shift.
On the flip side the British (at least in cruiser tanks) used very high-powered engines (or at least high torque) that allowed the drivers to concentrate of driving instead of shifting.
Soviets used the high powered V-12 but the transmission was so poor that many drivers rarely got into high gear, meaning that the book/proving ground speed was rarely reached.
Also means that the fuel consumption numbers were never right.
For those of us that have driven manual transmission cars, not using high gear is going to affect fuel milage even if the engine itself is good for low g/hp/hour
Ease of shifting is important. Back in the mid 70s I had the pleasure (not) of driving a fire truck with a poor engine swap and a crappy shift linkage. It had more power than the old engine and a lot more torque but it had a narrower rpm band, a crash box transmission (double clutching), a heavy flywheel, no accelerator pump on the carb and bent shifting rods in the transmission linkage. Truck also had manual (no power) steering which meant you had to do your shifting either before or after a turn, you did not have a 3rd hand to shift with while wresting the wheel.
If you blew a shift you had to stop the truck and start from 1st gear and standing start.

Perhaps there is another paper/study that concerns transmissions?

 

[Elan Vital]

Thank you.
what seems to be missing from these French calculations is the transmission set-ups.
You can get away with a lower powered engine IF you have more gears in the transmission. But that means a more expensive, larger, heavier transmission. You can also go too far in this direction as the Germans did. You also need to make sure such a transmission is easy to shift (at least somewhat) and that the drivers are trained to know when to shift.
On the flip side the British (at least in cruiser tanks) used very high-powered engines (or at least high torque) that allowed the drivers to concentrate of driving instead of shifting.
Soviets used the high powered V-12 but the transmission was so poor that many drivers rarely got into high gear, meaning that the book/proving ground speed was rarely reached.
Also means that the fuel consumption numbers were never right.
For those of us that have driven manual transmission cars, not using high gear is going to affect fuel milage even if the engine itself is good for low g/hp/hour
Ease of shifting is important. Back in the mid 70s I had the pleasure (not) of driving a fire truck with a poor engine swap and a crappy shift linkage. It had more power than the old engine and a lot more torque but it had a narrower rpm band, a crash box transmission (double clutching), a heavy flywheel, no accelerator pump on the carb and bent shifting rods in the transmission linkage. Truck also had manual (no power) steering which meant you had to do your shifting either before or after a turn, you did not have a 3rd hand to shift with while wresting the wheel.
If you blew a shift you had to stop the truck and start from 1st gear and standing start.

Perhaps there is another paper/study that concerns transmissions?

From the current knowledge of the archives, there is no overarching study on transmissions, but the ARL made a very detailed study with calculation notes on a mechanical transmission for the ARL heavy tank project of 1937 which was a 55-tonne vehicle powered by a 500 hp diesel engine meant to achieve a top speed of 30 kph, which was a challenging set of requirements compared to previous French tank projects. Link here: AA 206 4H2 41 - ARL mechanical transmission - Google Drive

The ARL came up with a fully-mechanical synchronized preselector gearbox with 5 forward speeds (up to a theoretical 32.6 kph) and 2 reverse speeds (3 kph theoretical max, the 1st reverse being used for very low-speed maneuvering) coupled to a steering unit giving 3 steering radii per gear. Gear change was assisted and all the necessary actions were combined so that the driver would only have to use a gear-change lever and a steering wheel. This is rather in line with the most convenient transmissions of the time period and WW2 for the driver.
However the heavy tank project was turned into the even heavier fortification assault tank program which mandated electric transmissions, and this transmission type doesn't appear to have been further developped for lighter vehicles.

---

The ARL V1 75mm SPG used a Cotal preselector gearbox with 8 forward speeds for up to 42.6 kph theoretical (42 at least achieved in trials) and one reverse speed, mated to a Cotal steering unit with 2 steering radii per gear. This is analogous to Wilson-style transmissions, but with electromagnetic instead of hydraulic clutches. This didn't give any trouble in trials and made driving very easy. The clutch was of an automatic centrifugal type which protected the mechanical elements from jolts, comparing favourably with the SOMUA 75mm SPG with a conventional clutch which suffered from greater wear. An extrapolated form of this transmission was used in the ARL 44 tank. Last I checked, the transmission was already able to withstand the torque and power of the intended 340 hp-class engine.

---

The S35 used a 5 forward 1 reverse speed sliding mesh gearbox (pretty much the regular mechanical gearbox of interwar/early war tanks) mated to a double differential steering unit which made steering convenient and was on the progressive side much like the Merritt-Brown, Wilson and Tiger transmissions. Only the final drives had to be changed to use the 230 hp uprated engine from the S40. It was suggested that the S40 use a Kegresse Autoserve gearbox which was a form of automatic gearbox which simplified driver training and actions, eliminated driving errors and reduced mechanical wear. The steering unit could also get hydraulic assistance to further reduce effort but the main goal was to reduce the frequency of maintenance compared to the original steering clutches. If the S40 indeed entered production with this system, it would have been perfectly easy to drive and steer, but no information is available on the maximum input power or durability.

---

The B1 Ter featured a different gearbox from the B1 Bis (with 5 forward speeds as far as we know, described simply as:

Gearbox studied by the ARL and built by the BAUDET-DONON-ROUSSEL company. This gearbox was designed for the torque of the 300hp engine and was 22cm shorter than the B1 Bis gearbox, allowing a larger crew compartment.
This gain was obtained by :
- Using two secondary shafts
- Mounting some shafts on needle bearings
- Replacing some bevel gears with globular screws/globoid worm gears

While the engine was driving the primary shaft still in gear, the first secondary shaft ended by a globular screw was used for the first, second and third forward speeds and the reverse gear (with an intermediary shaft), the second secondary shaft was used for the third and fifth forward speeds and drove the main differential via a bevel gear. The planetaries and satellites were of the spur gear type as it was easier to build than conical/helical/bevel gears. The auxiliary differential driving the steering mechanism was driven by a globular screw.

This gearbox was also cheaper and lighter than the B1 Bis' type. The needle bearings were eventually replaced by roller bearings as they were deemed responsible for some failures during trials. It was recommended to add a 6th forward speed. Without further detail, I can not comment on how it compared to other mechanical gearboxes of the era. A full description exists in the 4th report related to question n°270 "Char B experiments" in the archives, but I have not read it yet.

With the appearance of the new Renault V12 engine, a new gearbox was designed to operate beyond the maximum 2000 input rpm of the original type and for an input power of up to 400 hp (420 hp V12 minus losses before the gearbox). By April 1940, it was reported that a gearbox designed for an input power of 480 hp was designed for a "modified B1 Ter" which was possibly the future ARL battle tank, also named the B40.

The B40 gearbox was described long after the war by engineer Lavirotte (who oversaw Char B and other tank development in the late 30s and 1940) as "gearbox studied by chief engineer Martin-Prével (director of the ARL), mounted transversely, containing 6 forward speeds and 6 reverse speeds thanks to a reverser, a main differential, an auxiliary differential, both using spur gears driven by worm screws." I can not comment on this type either, other than it enabled yet another reduction in engine bay length.

Char B gearboxes were mated to the Naeder hydrostatic steering unit, which was described by German engineer Kniepkamp as giving the following advantages and disadvantages:

Drawbacks: heavier, bulkier and more man-hours intensive to make than other types, and the tank can't be steered at "Mittelpunkt" (in place?).

Advantages: can provide all steering radii continuously and without losses, which reduces/eliminates heat, smoke and noise generation when steering, reduces wear, has no delay, doesn't require much driving effort, doesn't jolt/stutter, ensures safe driving even at high speeds and the steering direction doesn't suddenly invert.

I will add that it had lower efficiency than a purely mechanical system, just like torque converters. The B1 Ter introduced a simplified Naeder with the mobile parts being fully interchangeable. The "modified B1 Ter" further simplified it by
- replacing its belt drive system by a gear drive system from the gearbox
- removing the drive system for the Henry pump (now driven directly by the gearbox) which was made smaller
- deleting the entire front end of the Naeder which removed the delicate play of the generator shaft, the ball bearings being replaced by double conical bearings, set once and for all by the manufacturer instead of having to be reset periodically by the crew or maintenance units
- Improving cooling oil flow using an internal turbine
- improving the seals (the Naeder had a tendency to leak oil, which also clogged the radiators)

The B1 Bis/ unmodified Ter also were to use a Naeder belt tensioner to increase the life of the belt.

Aside from the improved and simplified Naeder, the other two options were:
- a Poniatowski-type electric steering unit which had greater efficiency and eliminated steering efforts, the Chobert radiator and reduced maintenance but had some quirks which made driving less convenient
- a THAV hydraulic unit studied by Baudet-Donon-Roussel since late 1937 which was more compact and less complicated than the Naeder. It eliminated the belt drive from the Naeder in the unmodified B1 Ter, which was desirable as this belt drive was inefficient and insufficiently durable.

The B1 Ters also tested servo assistance before the Armistice. Overall the Char Bs can be described as using a regular manual gearbox with a steering system which is easy to use, with a trend towards providing assistance to the driver's controls, simplifying the steering steering system and making it more reliable, and increasing the maximum input power. The next step would have been to use a preselector, planetary or automatic gearbox coupled with a non-hydrostatic steering unit providing the same advantages as the Naeder without the same drawbacks.

---

The Renaults G1R and DAC 1 used a seemingly regular gearbox with a controlled differential with Lockheed assistance (so one of the worst forms of steering).

The AMX light tank used manual 5-speed gearbox coupled to a controlled differential as the absolute backup, but the baseline were two Wilson preselector units with geared steering (so quite good for WW2), and the desired final transmission was the Robin Van Roggen automatic type, which would be upscaled for 45-tonne tanks if found successful on the light tank.

---

The British were good with Merritt-Brown and conservative before that. The Tiger units were very strong and convenient to use but complicated to make. The Soviets saw the problem with their gearboxes in 1940 and started working on 8-speed gearboxes (KV-1S and IS series) and planetary turning mechanisms (IS-1/2). Op Barbarossa delayed or cancelled some of these developments (T-34 with 8-speed or planetary transmission).