Alternative strategy in the Battle of the Atlantic

Hull 440 (aka the Robert E. Peary) was built as a publicity stunt. Kaiser already had everything prefabricated to put the ship together. Construction began at midnight on Yard 2. Extra workers sped the assembly. Reporters had been brought in to witness the construction right from the start.

 

Over 1500 had severe defects in service. The vast majority of these were due to welding rather than rivetting in construction. What was determined was that the grade of steel being used in their construction was to brittle and that welding had caused additional local embrittlement along the welds. Because many of the welds were continious this allowed cracks to propagate along the hull for considerable lengths.
In addition, many fittings such as hatch commings hadn't been modified in shape (typically they still had square corners rather than rounded ones) that induced additional stresses at points like corners into the ship.

Two Liberty ships actually cracked in half as a result of this problem. On the follow on Victory ships the hulls were made less rigid, the metal chemistry and production process changed, and design features modified to prevent this problem.

 

I don't think the extra effort spent building the ships to be more torpedo resistant would have helped a great deal.
Of the 2,751 Liberty ships built in WW2, around 2,400 survived the war. The vast majority of the ships lost were pre-war built vessels.

 

The tendency of some the early Liberty ships to "crack" along the welded seams was eventually traced down by a group of researchers, formed in Washington, which studied the problem in detail and "fracture mechanics", as a discipline, was born. It was right after the first one actually broke in half, sailing between Alaska and Siberia, Feb. or April of 1943 that the Secretary of the Navy appointed a "Board to Investigate the Design and Methods of Construction of Welded Steel Merchant Vessels."

Turns out that of the approximately 5,000 welded merchant ships, of all sizes constructed during the war, about 1,000 actually suffered cracks appearing in different places on the ship, in varying degrees of severity. Some 190 ships (of various classes) sustained serious stress fractures, requiring immediate repair, but less than a dozen actually "broke in two", and all of the cracks appeared when used in frigid waters or freezing air temperatures. So twelve out of 5,000 "ain’t too shabby" when one considers the speed with which they were produced.

Originally the welders were supplied with a "rod grade" of 5010, which was a "match" for the rolled steel of the plates in tensile strength (50,000 psi), and using a paper flux coating. This remains a super easy rod to weld with even today, but not TOO strong however in a long seam joint.

The flux coatings were too high in both sulfur and phosphorous content, and the rod itself weakened the seam when exposed to extremes is temperature. This "mis-match" produced welding-induced embrittling of the hull and deck plates, with those cracks extending far beyond the norm due to a lack of any rivets (which had been present in previous designs).

The rivets would have normally interrupted any cracking paths in the seams. These failings were repaired with a different lower sulfur rod being supplied after 1943, and all the pre-'43 Liberty Ships being "retro-fitted" with a supporting band welded around the upper hull. No more of them broke in half after those repairs and changes in production material were made.

 

7018 and 6010 are the two most common welding grade rods today. I probably have 50 lbs of 7018 in the garage for the Lincoln right now.

 

In the early 1930s the japanese had a cruiser 'crack' around the hull. During its first regular training exercise in the North Pacific. The several causes were determined to be a combination of miscalculation of requirements for welding seams, and a general structural weakness in hull/frame design. Electrical welding was as new to Japanese shipyards as anyone elses in that era & they had their share of suprises.

The cruiser was the first of a projected class of four. The second, partially completed was extensively modified to correct the percived weaknesses. The third just laid down was canceled and the partially built hull scrapped, as were the materials stocked for starting the fourth. The 'US Naval Institute Proceedings' had a brief articall on this just a few years ago.

As for defeating German submarines... early and midwar decisions concerning ASW - particularly convoy and escourting, and providing aircraft for ASW had a lot to do with the tens of thousands civilian cargo ship crews who died in the North Atlantic.

"It never really was an issue during the war anyway, as neither German nor Japanese subs ever posed much of a threat to Allied maritime supply lines. According to Clay Blair, in "Hitler's U-boat War" Vol. 2, (page 708), of 657 total convoys on the critical North Atlantic convoy route between September, 1942, and May, 1945, comprised of 31,111 ships, just 247 (less than 1%) ships were lost. Those figures hold up for all periods and for the totals of all convoys in all areas. Less than 1% of ships transiting areas patrolled by Axis submarines were lost to submarine attack. So going to great lengths to protect merchant vessels against torpedo attack does not seem to make much sense. When actual losses are plotted against replacement tonnage it's clear neither German U-boats or Japanese subs ever had much of a chance of winning the "tonnage war"."

In terms of cargo sunk, or the total embarked for the UK the loss ammounted to around 4% - 7% annually. There was a surge to over 10% cargo lost in 1942 which is connected to Operation Drumbeat off the East Coast of the US and the related Carribian submarine operations. In 1943 when the Battle of The Atlantic' was at its crisis height loss of cargo embarked & destined for the UK started well below 10% at the start of the year and dropped to below 5% by midyear.

 

To what do you ascribe the difference between the percentage of ships sunk by U-boats and "cargo lost" on the UK run? Blair has done a pretty thorough job of researching and documenting the numbers of ships involved and those lost to enemy action, and I trust his numbers and methodology. It seems if you lose less than 1% of the ships carrying cargo to a given place, then the loss of cargo should follow pretty closely, unless of course, different methods, or different data bases, are used in accounting for the numbers.

 

Damm it. Shoot the sysop. I spent thirty minutes crafting a illustration of Ellis's and Costellos & Hughes charts, and lost it all due to the password bug in this web sites system.

To summarize: Ellis was counting only loss of 'displacement tons' destined for & docked in the UK. A different method of counting than Blair. A close look at the two shows that they more or less validate each other. Blair appears to be counting global losses. Do you read him that way?

The balance of what I wrote was a comparison of various loss analysis for 1939 - 1944 & particularly 1942 to 1943. What it shows it the real crisis of the Battle of the Atlantic was in 1942 vs the popularly suposed 1943. One cant blame the Allied leaders for thinking they were still in danger of losing in early 1943. While the situation was reversing from late 1942 they could not discount yet another Axis suprise. They had experinced way to many of those in previous years.

 

Welcome to the club! I spent hours putting together an analysis of global US troop shipments for 1942, and lost the whole shootin' match when the site decided I hadn't signed in.

Yes, that is the way it appeared to me. Blair did break Allied merchant ship losses down by area in the Summary section of Volume 2, but then combined all the sub-totals, by year, for all areas. His conclusions seem pretty consistent across the board for both numbers of ships, and tonnage. It is sometimes difficult to follow Blair, though when he converts the different types of tonnage measurements. Blair speaks of Gross Registered Tons or "GRT", but doesn't try to calculate the amount of cargo lost, only shipping.

That was what Blair seemed to be implying, that if there was indeed any "crisis" for the Allies, it was in 1942. Blair states that the Allies did not seem to realize that they had passed the crucial point in late summer, 1942, and that it would be all downhill for the Germans from the fall of 1942. He also states that Churchill's pronouncements on the U-boat danger should be viewed with a great deal of suspicion because he was trying to wring yet more aid form the Americans; most of what he claimed was exaggerated. For example, Churchill became alarmed when the UK's oil reserve fell from nine month's worth to seven months, and requested more tankers be put into service on the UK route to build it up again. But when the US began investigating, it was found that no other Allied region had more than about 10 weeks oil reserve, so nothing was done.

Blair says that the "Drumbeat" period saw some serious US losses, but even at the height of it, the Germans weren't sinking enough ships to have a chance of winning.

 

I'll have to accquire a copy of Blair I've been into his pages deep enough to know if he discusses the relationship of submarine losses and ASW aircraft. Whats your read?

As for the password glitch, I need to get backinto the habit of composing my posts on a WP document and pasting them into the post board here. having a copy saves a lot of grief. and anger

 

There are two volumes to "Hitler's U-Boat War"; the first volume covers the period from September, 1939 to August, 1942. and is sub-titled, "The Hunters". The second volume covers the war from August, 1942 through April, 1945, and is sub-titled, "The Hunted". I have Volume Oneand was able to find Volume Two in my local library. Blair does indeed discuss the different aspects of the impact of aircraft on the ASW campaign and U-boat losses due to aircraft, but mostly in Volume Two. Volume Two contains several lists of ASW aircraft assets assigned to the Atlantic regions at various times, and the losses Germany suffered from ASW aircraft. Volume One focuses more on surface ships, enigma, British and American mistakes, and the different philosophy each country held towards convoys and escorting. Blair pretty much shoots down the British nonsense about Admiral King being to blame for the relatively heavy losses in the first half of 1942. He also debunks several of the myths that have arisen, mostly due to sloppy scholarship, about the Battle of the Atlantic. Both volumes are well worth the price.

 

Interesting. What exactly does have to say about this?

 

Well, Blair points out that much of the criticism the British dump on King was based on incorrect or incorrectly understood information, poor scholarship and in some cases involving British historians, outright falsehoods. The British, for instance, disagreed with, and refused to accept the American policy in regard to the escorting of troop convoys. This led to King being blamed for the lack of escort availability for non-troop convoys. Blair suggests that some British historians used King as a scapegoat to distract attention from British mistakes in the Battle of the Atlantic. Blair doesn't claim King didn't make mistakes, but does assert that those he did make were not as significant as the British made out, and were in many cases exaggerated.

 

The only discussion of this I've seen with any depth focused on the German operation 'Drumbeat' and the US waters off the eastern seaboard and the Carribean.

To change the subject a bit, did Blair, or any others you read complain/comment on the possible disruption to the convoy protection caused by Op Torch and the follow on shipments to the Mediterranian? I've seen a couple claims of this & wonder if there is any substance to it?

 

Well, that seems to have been a major point of contention between the US and Britain, and "Drumbeat" was did result in a lot of ships lost simply because the US just didn't have enough escorts to effectively patrol the East Coast. But King was not the one to blame. For one thing, Roosevelt insisted on concentrating of building lots of smaller escorts which were not as effective as DE's and the larger Coast Guard cutters; King and other naval officers had opposed this but lost the argument with Roosevelt. I have seen some discussion of this issue in Friedman's Naval Design History series.

But just as problematical was the fact that the US had given away 60 escorts to the British and Canadian navies. Based on what the British historians have written about that deal, we shouldn't have done it. Those 60 ships would have made Drumbeat more like a whimper. Also coming into play was the pre-war turf fight between the USN and USAAF over who was responsible for coastal and offshore aerial ASW. The USAAF wanted to claim that job, but had never trained any aircrew for ASW duties.

Blair does discuss the Operation Torch period and mentions the effects of diverting escort resources to that operation. Blair points out that the British policy of rather casually convoying troop transports resulted in excessive losses of those types of ships. He indicates that the convoys from Norfolk to Gibraltar, especially the troop convoys, put a strain on the convoying system for a period. Ultimately, the US and Britain split responsibility for the escorting of the Atlantic convoys between the Northern route (Britain) and the Middle Atlantic route (US). Blair also indicates, in Volume Two, that the overall losses on the Middle Atlantic route were something on the order of .009%! As for the disruption, I got the impression it was more a matter of adjusting administrative responsibility than anything else; the ships were there, but the schedules and support functions had to be juggled. Of course, the finger-pointing did not change.

 

The best solution is to make faster merchantmen (17 to 20 knots ) which this tread states is impossible. Next best Idea turn every 50th ship into a CVE dedicated to convoy duty. And deploy liberator's sooner to close air gap mid-ocean.

Simply adding torpedo defences alone won't fly. Don't you need dedicated Damage control teams added to each merchantman? Where are the extra men coming from?

 

Increasing merchant ship speed was simply impossible. There wasn't enough capacity to build the steam turbines necessary. Britain lacked the diesel industry as an alternative and the US doesn't have the extra capacity in that field to devote to very large marine diesels beyond their already extant use. So, the only engine choice was the Vertical Triple Expansion (VTE) engine. While these could be built to give a ship as much as 20 knots of speed the trade off was in fuel use (range) and cargo load.
The US choice of the Liberty and Victory ships was a good one. These by then current standards were relatively fast (steaming at about 10 to 12 knots steadily). But, in a convoy the slowest ship sets the speed so even if you have fast ships it doesn't necessarily equate into faster sailing times.

The British did convert several tankers to carriers. There is photo attached below of one. The US opted for the CVE and built them in large number. But, it took about 15 to 18 months to build a CVE so there was a lag between start of the program and results.

Adding a damage control party to a merchant wouldn't have been that difficult. A typical team with mostly portable equipment would have say, 12 men plus use another 6 to 12 crewmen to supplement it. In fact, a party of this sort could have been moved from ship to ship to provide coverage with fewer crews. The big difficulty here is not manpower but equipment. The most difficult things to provide would have been the pumps and blowers,, particularly the pumps. These would have to have their own power (eg., gasoline or diesel motor) and have a useful fuel supply provided. Just making enough would take time too.
This was probably doable though.

 

As for increaseing speed I am not sure I would go so far as say impossible, but I totally agree that it would have been prohibitively costly to do so. So too would the cost to protect from torpedo damage. Brighter minds than mine have pointed out that high quality steel and construction methods used on warships would simply raise the per unit cost for each merchantman beyond acceptable limits. To compartmentalize enough to control flooding would greatly reduce cargo loads, needing more ships to move the same amount of cargo.

As for Damage control parties I will assume that the first figure of 12 men is additional crew, and that the 6 to 12 supplemental come from the regular crew compliment. That being said with a crew of 40 to 50 on a Liberty/Victory ship adding 12 men is a 25% increase in manpower per ship. With approximately 3,200 Liberty/Victory ships built that comes to 38,000 highly skilled seamen.

To be fair not all 3,200 were available at any given point, and not all of those that were available were at sea at the same time, so some switching between ships could have taken place. By the same token the 3,200 figure does not include all prewar merchantmen and any non Liberty/Victoy ships constructed during the war years. So it seems reasonable to conclude that somewhere in the area of 40.000 additional crew would be needed.

In WWII terms 40,000 does not seem high, but these would be highly trained crew akin to aircrew, or submariners not ordinary deckhands. Far more costly and time consuming to train and deploy and harder to replace when lost.

War like most things come down to cost versus benefit. I cannot see a economical way to make merchantmen any more survivorable in them selves. When the allies made enough VLR bombers and CVE's available for convoy duty, it then became impossible for the U-Boats to effectively strike at the convoys. Both measures could have been deployed sooner had the will been there to do so.

 

But, if all they are being trained in is basic damage control (firefighting and flooding) then the average training time would be about 30 to 45 days at most. They really don't need to know the ship's systems in detail sas the regular crew can provide those men when necessary. So, it would be possible to provide them.
But, the question would be if provided would they make a real difference? I thnk on merchant ships they probably wouldn't make that much difference simply because most of the time even a single torpedo hit is crippling damage. They might be able to stop progressive flooding and in some cases keep the ship afloat but in most they would likely just be additional casualties or survivors.

 

Considering they would be working in a high stress enviroment, a slightly longer training period might be needed, say 60 to 90 days. More importantly would be the amount of time spent on drill. Being civilians, it would depend upon who commanded the department, and the ship captains themselves. Likely ability would vary widely from ship to ship as opposed to warships. I suspect as you seem to that it would all be moot, as the ships design would likely work against them in any respect.