Earlier Type XXI - informed by Cold War developments

[TheMarcksPlan]

Tactically, the Type XXI is better than a Type IX for open ocean submarine warfare, but it isn't the claimed ten to thirteen times better.

Once again Mr. Gardner responds to quantitative analytical point (search area for submerged sub) with a non-analytical hunch. He doesn't even dispute the point, just raises factors irrelevant to this particular piece of the analysis.

It matters that it is not much harder to evade.

The idea that a 9-knot convoy can evade a 17-knot sub as well as a 7-knot sub is just facially absurd.

[TheMarcksPlan]

Another factor to discuss: T21's ability to kill its pursuers using "Gnat" homing torpedoes.

Older uboats had this capability as well (later in the war) and had some success in sinking escorts. The primary Allied countermeasure to the Gnat was the Foxer - a sound-producing decoy. Per RN analysis, the Foxer reduced the Gnat's success rate to ~15%.

In the context of T21, however, the Foxer has some serious problems:

• It can't be used above 14-15kn. So an escort chasing a T21 sprinting away from contact cannot use Foxer; the T21 can kill the escorts with Gnats. Foxer was a fairly large assembly system and there won't be time for an escort to deploy it if it detects a torpedo.
• It reduces maneuverability, especially when several escorts attempt to cooperate closely while all are towing Foxer buoys. As discussed upthread, multi-escort cooperation would have been essential for any chance of sinking a T21; such cooperation will be hampered or ruled out by Foxer.
• Its noise interferes with sonar and hydrophone location. Given the T21's higher speed and exponentially higher search area, escorts will need to be able to detect/track T21 at longer ranges. Foxer will hamper or rule or out this ability.

In addition, the T21 can fire 18 torpedoes in ~half an hour. Even with Foxer working, that's going to mean 3 dead escorts on average.

Finally, the Germans were developing solutions to Foxer. For example, a homing torpedo would run a semi-circular arc from the point of noise emission such that it would miss a Foxer while, if it homed on a ship's propeller, it would strike the side instead of the stern.

[Tom from Cornwall]

T21's underwater endurance figures are given on Wikipedia and elsewhere.

Thanks, although the figures you quoted seem to be the aspirational design speed/endurance as per this quote from the page:

This design promised an underwater speed of 18 knots for 1.5 hours or 12 to 14 knots for 10 hours with 4000 hp = 2942 kW. A speed of 5 knots was expected for 60 hours. In the course of the planning, the boat eventually grew to a size of 1600 tons.

Whereas actually achieved speed/endurance on main motors during trials is reported as:

Thus U 3507 reached 16.8 kn for 20 minutes and 16.5 kn for 50 minutes (admiralty constant C = 175) on 30 November 1944.

It's also worth noting this phrase:

It was calculated that with 5 kn of dipped march with schleichmotoren every 24 hours 3 hours battery charge with snorkel were sufficient to keep the charge level between 60 and 90%.

Unfortunately, the endurance statistics don't tell us what the battery capacity was at the beginning of the speed/endurance trial run and at the end of the run.

The Seraph had lower endurance because it was an older design.

Yes, it was an older design but it had gone through a large overhaul to allow it to approach the speed expected by the British of the Type XXI as described in the article I linked to:

The modifications included the uprating of the main motors from 1,400 to 1,600 shaft horsepower at full power, the fitting of coarser pitch T Class propellers that allowed the extra power to be converted to thrust, and the installation of a high capacity battery to extend her endurance. In addition, to reduce her drag Seraph was streamlined by fairing off apertures, such as, anchor holes, torpedo tubes and one-third of the free flood holes, as well as reducing the size of her hydroplanes. The forward planes were also fixed in the "out" position and given a more powerful control mechanism. The forward periscope, radar mast, Oerlikon anti-aircraft guns and the 3" deck gun were removed, and finally the profile of the conning tower was reduced.

Obviously, the Type XXI performance was impressive for the period and if the Germans had been able to deploy large numbers of Type XXI's much earlier in the war there would have been larger Allied shipping losses.

So an escort chasing a T21 sprinting away from contact cannot use Foxer; the T21 can kill the escorts with Gnats.

Really? Did the Gnat not need to be fired in the general direction of a target? Could the Gnat be fired at the submarine's max speed and then loop back to attack a following escort?

The idea that a 9-knot convoy can evade a 17-knot sub as well as a 7-knot sub is just facially absurd.

"facially absurd" - does that mean one of these -

If the Allies could still read German U-boat traffic then wouldn't evasive routing of a supposed 17 knot submarine in the Atlantic Ocean be just as effective as evasive routing away from a 7 knot submarine? You do realise how big the north Atlantic is and how small the detection range of the Type XXI was?

Regards

Tom

[Takao]

Another factor to discuss: T21's ability to kill its pursuers using "Gnat" homing torpedoes.

Older uboats had this capability as well (later in the war) and had some success in sinking escorts. The primary Allied countermeasure to the Gnat was the Foxer - a sound-producing decoy. Per RN analysis, the Foxer reduced the Gnat's success rate to ~15%.

In the context of T21, however, the Foxer has some serious problems:

• It can't be used above 14-15kn. So an escort chasing a T21 sprinting away from contact cannot use Foxer; the T21 can kill the escorts with Gnats. Foxer was a fairly large assembly system and there won't be time for an escort to deploy it if it detects a torpedo.
• It reduces maneuverability, especially when several escorts attempt to cooperate closely while all are towing Foxer buoys. As discussed upthread, multi-escort cooperation would have been essential for any chance of sinking a T21; such cooperation will be hampered or ruled out by Foxer.
• Its noise interferes with sonar and hydrophone location. Given the T21's higher speed and exponentially higher search area, escorts will need to be able to detect/track T21 at longer ranges. Foxer will hamper or rule or out this ability.

In addition, the T21 can fire 18 torpedoes in ~half an hour. Even with Foxer working, that's going to mean 3 dead escorts on average.

Finally, the Germans were developing solutions to Foxer. For example, a homing torpedo would run a semi-circular arc from the point of noise emission such that it would miss a Foxer while, if it homed on a ship's propeller, it would strike the side instead of the stern.

More fantasy from TMP...

Actually, it was less than 15%...over 700 firings for 77 sinkings is, at best 11%.

This is also quite contradictory to the RN author you previously quoted as saying the Type XXI would have near 100% success rate with acoustic torpedoes & pattern running torpedoes.


Problems with your list...
1. If the escort is pursuing the Type XXI - How does the Type XXI launch a torpedo at it's pursuers with no aft facing torpedo tubes? If the Type XXI does launch an acoustic from it's fore tubes, and the torpedo is lucky enough to complete the 180 degree course change...Then what is the loudest thing between the torpedo & escort...The Type XXI travelling at top speed...So what will the acoustic home in on? The Type XXI.

Noise interferes more so with passive sonar...Not so much with active sonar.

Further, Foxer was towed aft, behind the screws, so it would be in the ships dead zone any way. Also, if passive sonar worked the way you think it does...No ship would be able to detect anything because of their own noise, however passive sonar does not work that way.

The circular run of an acoustic torpedo is to reacquire it's target, not to hit the ship in the bow. After all, the torpedo does not know if it missed left or missed right. Modern homing torpedoes do this today if they miss, they will try to reacquire the target.

[Takao]

I presume your source is Eberhard Rossler's book
"The U-Boat: the Evolution and Technical History of German Submarines".

T21's underwater endurance figures are given on Wikipedia and elsewhere.

Thanks, although the figures you quoted seem to be the aspirational design speed/endurance as per this quote from the page:

This design promised an underwater speed of 18 knots for 1.5 hours or 12 to 14 knots for 10 hours with 4000 hp = 2942 kW. A speed of 5 knots was expected for 60 hours. In the course of the planning, the boat eventually grew to a size of 1600 tons.

Whereas actually achieved speed/endurance on main motors during trials is reported as:

Thus U 3507 reached 16.8 kn for 20 minutes and 16.5 kn for 50 minutes (admiralty constant C = 175) on 30 November 1944.

It's also worth noting this phrase:

It was calculated that with 5 kn of dipped march with schleichmotoren every 24 hours 3 hours battery charge with snorkel were sufficient to keep the charge level between 60 and 90%.

You should have added the previous few sentences in your quote that explains the discrepancy. To summarize, the originally intended motors were to have provided 3,500Kw of power, whereas the motors installed in the Type XXI class only provided 3,100Kw of power. Thus, when U-3507 achieved her sprint speed of 17.2 knots on November 21, 1944, she would have been able to make or surpass 18 knots with 3,500Kw motors & 2/3rds reduction of limber holes/flooding slits.

Unfortunately, the endurance statistics don't tell us what the battery capacity was at the beginning of the speed/endurance trial run and at the end of the run.

If you look at the previous page, it gives you an idea though

The test bed results for the large main motors had shown a maximum 1,800kW performance at 5,500A current-intake, and engine revolutions of 1,675rpm. With fully charged batteries, this performance could be maintained for one hour. Normal boast practice, however, was that each double motor would attain only 1,650kW for 20 minutes, or 1,550kW (= 2,100hp) for 50 minutes, because batteries, even when fully charged, did not hold their intended 360 volts charge at maximum performance. A test in U3506 showed that with batteries carrying a charge of 361 volts and a constant discharge of 2 x 5,540 Amps, the charge in the electric motors not supplying power was 350 volts, and when supplying power, only 336 volts. After thirty minutes, the charge was down to 332 volts, after 50 minutes, discharge was down to 305 volts. After that it went down rapidly.

[TheMarcksPlan]

Did the Gnat not need to be fired in the general direction of a target? Could the Gnat be fired at the submarine's max speed and then loop back to attack a following escort?

Not sure on the answer to that specific question. I know the Germans had a torpedo that was "position independent" - i.e. it could be programmed to run a pattern regardless of its initial bearing so could be fired forward against targets to the sub's rear. I don't know whether they combined that capability with a homing torpedo that would, presumably, have delayed activation of the homing function if position-independent. Doesn't seem like a terribly difficult technical problem once you've solved position-independence. Just add a timer for the homing function. Does anyone know?

On the tactical point you're misunderstanding the scenario: T21 doesn't have to fire at escorts while fleeing. Any decent commander will have situational awareness of escort location when he attacks; he can fire a salvo at the convoy, reload in 5 minutes, and then turn on the approaching escorts. Or if the escorts are running Foxer, he knows he can outrun them. The point is the escorts can't both (1) run Foxer and (2) catch a fleeing T21. They have to choose whether to be vulnerable to a dozen homing torpedoes or whether to have any chance of catching a fleeing T21.

If the Allies could still read German U-boat traffic then wouldn't evasive routing of a supposed 17 knot submarine in the Atlantic Ocean be just as effective as evasive routing away from a 7 knot submarine?

It's not about Ultra, it's about speed. Of the 1,000+ Uboats Germany produced, only ~300 did any damage. Most just saw ships going by over the horizon. A T7/9 that can't surface needs the convoy to run it over because otherwise it can't catch it - thus the focus on NW Approaches inshore in 1944.

"facially absurd" - does that mean one of these

It'd be like saying the Sopwith Camel and Spitfire have similar interception abilities against WW2 bombers. Just plainly absurd, plainly motivated by a desire to maintain a pre-set view when confronting facts and analysis.

[TheMarcksPlan]

More from RNASW49, p103-104:

Had the enemy been able to employ these submarines in the manner proposed,
Ashbourne concluded, they would have been able to defeat the most effective
British counter-measures and, since the Type XXI did not need to use the surface,
except to schnorkel for short periods, aircraft would be virtually powerless to sink
them while in transit, except on rare occasions. The Type XXI’s weakness was
reconnaissance, so support by aircraft would be especially valuable in locating
targets. An advantage was the use of the ‘Squash’ or ‘Kurier’, pulsed radio system,
which allowed short, formatted messages to be cleared in less than one half of a
second. Existing ship-borne direction finding equipment could not exploit these
signals, though this might be possible from shore stations within the next few
years.

The Ashbourne quoted is Lord Captain Ashbourne, an eventual Admiral in the RN who had extensive war-time submarine experience. https://www.uboat.net/allies/commanders/3070.html

But hey - all who want to accept the hunches and other non-analytical treatments offered against T21 in this thread - as opposed to the opinions of Ashbourne and others - will surely continue to do so.

-----------------------

The second part of the quote addresses the non-tactical problems that I have so far not discussed in depth. For purposes of conceptual clarity, I'm first focusing on what happens when a T21 encounters a convoy, then moving to how often those encounters would happen.

Preview: If the T21 is, say, 5x as effective per encounter as T7/9 were in 1942, then 1/5 the "search" effectiveness per boat is required to achieve the same sinkings per boat-day-at-sea.

And of course if attrition rates are lower for T21 then "boat-days-at-sea" would be rising at a rate greater than for T7/9 commissioned at similar numbers.

I don't expect most members to follow this kind of analytical path - I expect conceptually incoherent swings between the different factors of analysis - but that's how I'm going to proceed.

[Takao]

This is the Ashbourne that thought the Type XXI would achieve near 100% success rate with pattern-runners & homing torpedoes?

Despite current information & analysis showing a success rate far lower (11% for homers alone).

Curious as to why his opinion/hunch/non-analytical treatment should be taken as gospel, when it is so off?

Oh...Yeah...It agrees with your hypothesis.

[TheMarcksPlan]

An important distinction is necessary here between:

(1) Pattern-running torpedoes and

(2) Homing torpedoes

A pattern-running torpedo runs back and forth within a pre-set grid (i.e. a convoy) until it hits something or runs out of fuel. This is what the Germans and RN predicted would have ~95% success rate.

A homing torpedo uses passive hydrophone effect to lock onto a noise source - here a propeller. This has a lower success rate than a pattern running torpedo, except against faster and maneuvering targets.

Apologies to any newcomers to WW2 history who were confused by this terminology. ASW can be complex subject and it's easy to make these mistakes, especially when one is new to the topic.

Also there were separate people in the RN named Mcrea and Ashbourne. Each adjudged the T21 unmanageable by WW2 escorts, each is cited in RNASW49, yet each is a separate person.

[TheMarcksPlan]

Obviously, the Type XXI performance was impressive for the period

Not really.

It had an inefficient pressure hull design - "8"-shaped instead of circular to save time when adapting it from the Walther boat designs.

It could have used bigger diesels to reduce battery charging time.

It was still smaller than American fleet submarines; larger size would have enabled greater time on station and slightly greater speed at constant HP/ton.

(I'm sure others will be happy to list other flaws in the T21)

T21 was a decent implementation of an obvious idea: make subs faster underwater.

This obvious idea wasn't needed by other countries, which either didn't engage in large-scale sub warfare or were doing fine with the old philosophy (USN).

-------------------------

The essential insight, IMO, is that one doesn't have to be good at submarine warfare to inflict immensely asymmetric costs on the enemy.

Sub warfare is therefore the marine analogue of guerilla warfare: You don't have to possess any exceptional military effectiveness to make guerilla warfare extremely and asymmetrically expensive.

I see 1943-45 as brief blip in which ASW was able completely to master the submarine; had Germany simply maintained its earlier so-so performance it would have had significant strategic implications.

The T21 would have far exceeded maintaining earlier German performance, at least initially. Even if countered more effectively by later technological developments, it would have maintained the asymmetry of expense in BoA.

[TheMarcksPlan]

Re the issue of faster convoys, from War Production Board minutes of May 25, 1943:

Mr. Gibbs stated that the cost of construction per dead weight ton for the various types of ships reflects the relative consumption of materials and man-hours of labor and he presented the following data: Relative cost per deadweight ton

EC-2 (Liberty) ............. $170
Victory . . . . . . . . . . . . . . . . . . . . . . 238
C-2 . . . . . . . . . . . . . . . . . . . . . . . . . 313

The Vice President inquired whether the Victory ship is not better than the Lib erty ship for war purposes. Mr. Gibbs stated that according to the principles he had outlined the Victory ship is better than the C-2 but not as good as the Liberty ship. The Liberty ship has a lower production cost than the Victory ship in terms of materials and manpower, and the advantage of the latter's higher speed cannot be realized until long after its production has started and a sufficient number are in operation to utilize their potential speed in convoy.

As I say in the OP, a predictable and somewhat-effective Allied response to T21 would be faster merchant ships.

As the WPD recognized, however, this response would be both enormously expensive and would take a long time.

The RN analysis presented in this thread has conclusively shown that the convoys and escorts prevailing in WW2 would have been helpless against T21.

To replace those systems with faster, higher-tech convoys/escorts would have taken at least a few years and would have required the diversion of Allied resources from other fields.

[Lars]

If the XXI is employed earlier, expect the strategic air war to shift focus at least from the British side. The play would be to mine the XXI-ports, mine the Baltic training grounds, bombs the waterways which transported the XXI-sections, and bomb the yards which assembled the XXI-boats. Some XXI-boats would of course always get assembled and some training would always take place so it would only slow and not halt the XXI-production. And of course the shipyard cities would be beefed up flak-wise and the night-fighters would know of the special Allied interest in the shipyard cities. Area bombing of German cities would take second place but only after a Churchill-Bomber Harris row.

If the USAAF joins in, the bonus on the German side could be great especially if it means that the USAAF scales down their oil campaign from May 1944 onwards.

[Terry Duncan]

More from RNASW49, p103-104:

Had the enemy been able to employ these submarines in the manner proposed,
Ashbourne concluded, they would have been able to defeat the most effective
British counter-measures and, since the Type XXI did not need to use the surface,
except to schnorkel for short periods, aircraft would be virtually powerless to sink
them while in transit, except on rare occasions. The Type XXI’s weakness was
reconnaissance, so support by aircraft would be especially valuable in locating
targets. An advantage was the use of the ‘Squash’ or ‘Kurier’, pulsed radio system,
which allowed short, formatted messages to be cleared in less than one half of a
second. Existing ship-borne direction finding equipment could not exploit these
signals, though this might be possible from shore stations within the next few
years.

The Ashbourne quoted is Lord Captain Ashbourne, an eventual Admiral in the RN who had extensive war-time submarine experience. https://www.uboat.net/allies/commanders/3070.html

But hey - all who want to accept the hunches and other non-analytical treatments offered against T21 in this thread - as opposed to the opinions of Ashbourne and others - will surely continue to do so.

-----------------------

The second part of the quote addresses the non-tactical problems that I have so far not discussed in depth. For purposes of conceptual clarity, I'm first focusing on what happens when a T21 encounters a convoy, then moving to how often those encounters would happen.

Preview: If the T21 is, say, 5x as effective per encounter as T7/9 were in 1942, then 1/5 the "search" effectiveness per boat is required to achieve the same sinkings per boat-day-at-sea.

And of course if attrition rates are lower for T21 then "boat-days-at-sea" would be rising at a rate greater than for T7/9 commissioned at similar numbers.

I don't expect most members to follow this kind of analytical path - I expect conceptually incoherent swings between the different factors of analysis - but that's how I'm going to proceed.

Your low opinion of other members here is probably something you should keep to yourself, it is unlikely to make people receptive to your ideas when you refer to them in such a manner.

With regards to the Type XXI, how does Germany actually build them? They were designed to take advantage of modular construction from dispersed industry in order to make the production difficult to disrupt. In reality, when the parts turned up they didn't fit together too well. Maybe this isn't critical in a surface ship, but it is for a submarine as pressure on the hull dictates a need for uniform strength. From the parts arriving at the assembly yards a few boats were built, of which even fewer were seaworthy. The production method was flawed and not too likely to make good use of resources. This could have been avoided by returning to the traditional methods of construction, but then that comes with the same production schedule that Speer was attempting to shortcut. These issues were never really resolved. A 4% success rate in actually building the things from the parts available is laughable.

However, if we assume the supply of these boats is possible it is not going to happen early on in the war due to the design process taking years to arrive at the Type XXI, they did not spring from thin air, they are the product of war experience and so you need actual time at war before they even become possible, somewhere like the end of 1942 just for the idea!

The high submerged speed is useful, but it also means the ship is running blind as it cannot detect anything due to the noise of its own engines unless it slows down to speeds where the escorts are now a threat. The escorts do not need to sink the submarine, only to prevent it from getting into a position where it can launch an attack. As others have pointed out, the convoys may even be routed out of the way of the Type XXI or it simply doesn't get lucky and misses contact. The passages from Ashbourne are interesting, especially the use of the word 'Had' at the start of the assessment, implying it was beyond the capabilities of the time and therefore examining a worst-case scenario rather than a realistic one.

Maybe if you allow a 1947-8 Germany to fight a 1943-4 Allied force then yes, they may make a small difference, but only for the time it takes for an Allied response to them. If I were to suggest allowing a 1945 Allied escort group to defend against a typical 1939-40 German wolfpack attack in order to prove submarine warfare was a losing game you would rightly say I was not being balanced or realistic, and the same applies in reverse when people are presented a 'what if Germany perfects everything and the Allies have to stick with what they had historically' type scenario.

[TheMarcksPlan]

If the XXI is employed earlier, expect the strategic air war to shift focus at least from the British side. The play would be to mine the XXI-ports, mine the Baltic training grounds, bombs the waterways which transported the XXI-sections, and bomb the yards which assembled the XXI-boats. Some XXI-boats would of course always get assembled and some training would always take place so it would only slow and not halt the XXI-production. And of course the shipyard cities would be beefed up flak-wise and the night-fighters would know of the special Allied interest in the shipyard cities. Area bombing of German cities would take second place but only after a Churchill-Bomber Harris row.

If the USAAF joins in, the bonus on the German side could be great especially if it means that the USAAF scales down their oil campaign from May 1944 onwards.

This is right but it's not all that different from OTL during '42-'43. From O'Brien's How the War was Won, Chapter 7:

The other area where the British and the Americans ended up
employing extra resources to combat the German submarine threat in
1942 and 1943 was in the growing strategic bombing campaign.
Within this expensive campaign that was being waged against
Germany, anti-submarine targets received a high priority. For the first
six months of 1942, the USAAF was devoted entirely to different
anti-submarine operations

During 1943 as a whole, approximately half of the bombs dropped by
the USAAF based in Britain were on German submarine targets.98 The
figure for the RAF was approximately 20 percent

Germany produced more subs in 1943 than in any other year. To double bombing efforts against subs would require 100% AAF focus or diversion of RAF resources. As RAF wasn't good at precision bombing, its use is inefficient.

From '44 the Allies have more ability to devastate U-boat production but the opportunity cost is very high.

[TheMarcksPlan]

However, if we assume the supply of these boats is possible it is not going to happen early on in the war due to the design process taking years to arrive at the Type XXI, they did not spring from thin air, they are the product of war experience and so you need actual time at war before they even become possible, somewhere like the end of 1942 just for the idea!

Were the KM competently invested in submarine R&D, they likely would have come up with the idea earlier. The motive was already there: Donitz had briefly to cease all night-time surface attacks in May 1941 due to radar. That's 1.5 years earlier than OTL T21 genesis.

Within the KM, it took inordinately long for the Walther types to get strategic-level discussions and for KM engineers to be invited to discuss the proposal. When this finally happened in Fall '42, one of the engineers made the obvious point that maybe we should just fill the extra hull space with more batteries instead of hydrogen-peroxide.

All that has to happen is that some competent, higher-level engineers/designers see the Walther proposals earlier and the obvious idea would have taken hold.

Donitz bears a lot of the blame here as well. He made basically zero efforts from '38 onwards to invest in submarine design.

As with all things, a strategically-better Hitler would have made a difference also. Had he stuck to his original anti-big-fleet position and refused Raeder's Plan Z scheme and its strategic drift, the KM would have focused earlier and better on submarine development.