Monday, September 30, 2013

Mister David A.Salmon - The State Department’s master of ciphers

So far I’ve presented lots of interesting information on the compromise of the State Departments strip cipher during WWII.

The M-138-A strip cipher carried the most important diplomatic traffic of the United States (at least until late 1944) and by reading these messages the Axis powers gained insights into global US policy.
The strip cipher was not a weak system cryptologically, even though it could not offer the security of cipher machines. The success of German and Finnish codebreakers was facilitated in many cases by the poor way that the system was used by the State Department.

Each embassy was provided with 50 ‘special’ alphabet strips and 50 ‘circulars’. The ‘specials’ were used for direct communications between that embassy and Washington. The ‘circulars’ were used for communications between embassies and for messages sent from Washington to more than one embassy.
Each day 30 different strips were selected from the 50 and entered into the metal frame. The strips and the order that they were inserted in was the daily ‘key’.

The major mistakes made by the State’s cipher department were:
1). During the period that the strips were valid there were only 40 different rearrangements for them and not a separate key for each day.

2). The ‘special’ strips were not destroyed after the period of validity had passed but instead in many cases they were sent to another embassy to be used there.
3). The ‘special’ strips used by each embassy differed from those used at other embassies but it seems that the daily ‘key’ was the same for all. (may have changed during the war)

4). The same messages were sometimes sent in ‘circular’ and ‘special’ strips.
5). The same messages were sometimes sent in low/mid level book codes and the strip cipher.

Without these mistakes the Axis effort would have been considerably hindered.
What kind of department was responsible for the way that the State Department used the M-138 during the war? This question required quite a lot of research into the archives in order to track down the relevant documents.

There is a ‘History of the Division of Cryptography’ which can be found in NARA collection RG 59.

 




According to this study (and other documents) prior to September 1944, when the Division of Cryptography was established, the State Departments cryptographic functions were performed by a small unit in the Division of Communications and Records, under the supervision of the Division chief mr David A.Salmon. When in August ’43 he was assigned Assistant Security Officer in the Office of the Assistant Secretary for Administration he retained his cryptographic responsibilities.

 
Mr Salmon’s small unit was not a cryptologic department and could only handle the basic tasks of preparing cipher material and the instructions on their use. Obviously this was one of the reasons that there were serious problems with cipher security during the war.

 


However it would be dishonest to try to pin all the blame on mr Salmon and his unit. Apparently the State’s leadership had rejected the purchase of secure cipher machines in 1941 due to the financial cost. Instead they kept using the same codebooks (Gray, A1,B1,C1,D1) for decades.  The Brown code and the Strip cipher were the only new systems introduced in the 1930’s.
 
Anything else to report? Well actually it seems that mr Salmon may have been a Soviet spy all along! According to the wikipedia page on mr Salmon the book ‘Spies: The Rise and Fall of the KGB in America’ names him as a Soviet spy…

Thursday, September 26, 2013

Another nail in the T-34’s coffin

The well known Soviet T-34 tank has been called the best tank of WWII in countless history books. I have presented a very different picture of its capabilities and performance here.

I’ve also written about the experiences of a German unit that used the T-34 model ’43 in combat and also presented information on the German response to the KV and T-34 tanks.
In order to find more information on various aspects of WWII I always have a look online at various sites and forums. Sometimes one can find an interesting link or a quote from a book that reveals new information. Unfortunately most of the time I’m left shaking my head at the moronic arguments, circular reasoning and lack of common sense that one often finds online.

However with a bit of luck some diamonds do turn up in the most unlikely places. In the comments of a piece at the world of tanks site ‘For the record’ the commenter ‘ Mo’ (comment of September 6, 2013 at 1:17 am) linked to a US study of Soviet equipment that doesn’t seem to paint the Soviet equipment in such a good light:

Good quality WWII Russian armor and shell steel? The CIA disagrees.
Check out another CIA analysis of Russian armor and welding.
http://www.dtic.mil/cgi-bin/GetTRDoc?AD=AD011426
It must be an unlucky coincidence that his link lo longer works.

The report ‘Review of Soviet ordnance metallurgy’ is dated 10 April 1953 and can be downloaded from here.
 
It has interesting details on the quality (or lack thereof) of Soviet weaponry. I’m particularly interested in the T-34 evaluation so here are some quotes:

‘The ordnance Corps's first contact with modern Soviet tank armor was in 1943 when two tanks were provided to this country by the Soviet government for performance tests at Aberdeen Proving Ground. These tanks were the T-34 medium tank and the KV-1 heavy tank.’
Note: The T-34 sent to the US for testing belonged to a special batch built in spring 1942 at Nizhny Tagil. Five T-34 were built with one being sent to the USA, one to the UK, two to the front and the last remaining at the plant.

The plant was specifically chosen because it had the highest quality of T-34 production, at that time. All components were built with the outmost attention to quality. As such the tanks were not indicative of the average production but were of much higher quality. Source: ‘Tankovy udar. Sovetskie tanki v boyakh. 1942-1943’

What was the US assessment? Let’s see:
‘The armor components of the T-34 tank, with the exception of the bow casting which was unheat-treated, were heat-treated to very high hardnesses (430-500 Brinell), probably in an attempt to secure maximum resistance to penetration by certain classes of armor-piercing projectiles even at the expense of structural integrity under ballistic attack.’

‘The quality of the armor steels ranged from poor to excellent. Wide variations in production technique were indicated; some rolled armor components were well cross-rolled while others were virtually straightaway rolled………The bow casting of the T-34 tank was very unsound and would have been rejected under American standards.’
'The design of the welded joints was characterized by dovetailing such that the edges of the lighter plates were set into niches machined or flame-cut into the heavier sections so that the surfaces of the lighter plates were approximately flush with the edges of the heavier sections…..Although the fundamental design of the joints appeared excellent, the fit-up, appearance, and execution of the joint design and welding was generally poor.’

Shallow penetration, poor fusion, severe undercutting, porosity, and cracking was observed in most of the welds and probably resulted from improper manipulation of electrodes which might not have had suitable operating characteristics….. These obvious defects, together with low strength and pour metallurgical structure of ferritic weld deposits, indicate that the welded joints would have poor resistance to severe shock.’
Now I know what you’re thinking. It was 1942 so the quality problems were undeniably due to the war situation and the relocation of industry to the Urals. Obviously the US report will mention this:

‘The results obtained from the metallurgical examination of these early world war ii Soviet tanks have been described in some detail since they are exactly the same as have been obtained from all examinations performed since then of Soviet tanks which were recovered in Germany after the end of world war ii, and on Soviet tanks which were captured in Korea during 1950-52. The Ordnance Corps has examined several Soviet JS-II which were found in Germany and several Soviet T-34 tanks from both Germany and Korea.’
Hmm I guessed wrong…. No worries let’s continue with the quotes:

‘Some of the armor steels have surprisingly high toughness considering the very high hardness levels but many of the armor steels, even the softer ones, are very brittle.’
‘The very high hardness encountered in most Soviet tank armor has caused much unnecessary concern regarding the relative ballistic performance of the hard Soviet armor and the softer American armor. Many people associate high hardness with high resistance to penetration. Although this is true, within limits, in the case of attack of armor by undermatching projectiles (i.e. caliber of shot is less than the tnickness of the armor) particularly at low obliquities of attack, it definitely not true when the armor is attacked by larger caliber shot at higher obliquities of impact’

You don’t say…. So maybe the T-34’s 45mm hull armor was not the best choice given the widespread use of the German KwK 40/Pak 40 75mm gun? (used in Pz IV, Stug and self-propelled vehicles). Who would have thought?
‘Although welds in Soviet tanks are inferior in quality and much more brittle than corresponding, welds in American tanks, this condition has not been a major factor in impairing the battlefield performance of  Soviet armor. Poor joint fits, sloppy appearance, jagged and rough finishes should not divert attention from the fact that the Soviet tanks are rugged and battleworth and require many fewer man-hours of labor and precision machine tools, jigs, and fixtures to construct than American tanks of corresponding offensive capabilities.

This is a very interesting argument. Quantity over quality. But notice that there are no numbers to back it up. The author simply assumes that poor construction means a Soviet emphasis on production and not an inability of Soviet industry to produce quality products.
This is confirmed in the next sentence:

‘it would be very interesting to compare, for example, the relative man-hours of labor and investment in machine tools to construct equivalent numbers of the American 76 mm, Gun Tank T41 and the Soviet T-34/85.’
Yes that would be interesting but the author hasn’t done it, probably due to lack of reliable Soviet data…

The study concludes:
‘It must be borne in mind that the Soviet ordnance materiel described in this paper was mostly of world War II manufacture and represents design concepts which, for the greater part, were established as early as 1940-1942. It cannot be said with any certainty that these design concepts are, in all cases, still adhered to by the Soviets.’

‘From  a metallurgical point of view, it would appear that the Soviets have attained equality with this country in the matter of technical information but not in technological development or in skill and training of metals workers such as weldors, foundrymen and machinists.’
There is also a restatement of the quantity over quality argument, again however with no data to back it up.

‘In closing, it should be emphasized that this country could do well to emulate the Soviet practice of employing finely machined finishes only where needed. The same applies to high quality, carefully prepared welded joints, castings, and other metal products. Detailed attention to aesthetic appearances is costly, time consuming, and, throughout the history of man, is not known to have won a single war.’
My advice is to read this study if you’re interested in Soviet WWII equipment.

Tuesday, September 24, 2013

Starving Britain into submission – a flawed strategy?

During WWII the Germans tried to take advantage of the fact that Britain was dependent on imports of raw materials and foodstuffs for its economy to function.

The U-boats tried to sink as many merchant ships as possible so that their valuable cargo would be lost to the British.
The article Fighting fit: how dietitians tested if Britain would be starved into defeat reveals that the British authorities carried out an experiment in order to discover what would happen if the populace had to depend only on homegrown foodstuffs. Would they starve to death?

According to the article:
British food production in 1938 became the basis for the experimental diet: one egg a week (a third of the pre-war consumption); a quarter of a pint of milk a day (half the pre-war consumption); a pound of meat and 4oz of fish per week, assuming trawlers would be commandeered for patrols. No butter and just 4oz of margarine. But they could eat as much potato, vegetables, and wholemeal bread as they wanted. The eight guinea pigs would follow this diet for three months.

Surprisingly the participants in the experiment did not suffer from malnutrition but only from less serious ‘side effects’:

Happily, the gloomy spectres of famine oedema, scurvy, and anaemia did not arise. The guinea pigs felt fit and well on the ration and could do their usual work. But there were two main difficulties. One was that meals took a long time to eat. Wholemeal bread without butter took ages to chew. The sheer quantity of potato needed to make up calories also took time to eat. All the fibre in the diet caused 250% bigger poos. They measured it. The other problem with eating all that starch was the amount of flatus – gas – that it produced

Monday, September 23, 2013

NSA spying and the threat of terrorism/spying/cyber war etc etc

Ever since the former NSA employee Edward Snowden revealed the extent of NSA’s and GCHQ’s internet spying there has been a backlash against the secretive and obviously unlawful operations of NSA and its allies.

On the one hand people have complained about the indiscriminate interception of the entire world’s internet and phone traffic, while the other (much less numerous) side, made up of people associated with the NSA and the US intelligence community, has tried to make the argument that even if certain laws were broken it was all in the interest of ‘national security’.
According to their side people should just shut up and deal with the complex realities of cyber warfare, internet spying and all that jazz. Oh and of course we shouldn’t listen to Snowden cause he’s just a Chinese/Russian spy and has psychological problems and and and.

That strategy was more or less effective at the start of this story and I remember that many (independent?) media started focusing on Snowden and not on the Orwellian policies of the NSA.
It is a testament to the professionalism of Snowden’s collaborator Glenn Greenwald that important material is released in a steady basis, so the media aren’t overwhelmed by the information. This means that critics have to focus on the NSA activities and cannot sidetrack the discussion with accusations about Snowden’s motives or his personal life.

This strategy of the Snowden team has left the ‘defenders of the realm’ holding their dick in their hands (as we say in Greece) 
Now the question of surveillance/spying and the limits that have to be imposed is a difficult issue. The defenders of the NSA can claim that they need to intercept everything, subvert codes and break computer software because that will help them arrest spies, terrorists and other bad guys.

Looking back through history it is interesting to compare their efforts with the activities of the British intelligence agencies during WWII.
The Brits had to deal with foreign states like Germany, Japan, Italy and the Soviet Union that had extensive espionage networks throughout the world and were often supported by other ‘neutral’ countries.

Although in 1939-40 British intelligence was woefully inadequate during the war their performance picked up and they were able to dismantle enemy networks and build up their presence in ‘neutral’ countries like Spain, Turkey, Sweden and Switzerland.
In their efforts they were assisted by signals intelligence. The German intelligence agency Abwehr used the Enigma G cipher machine for communication between main stations. This device was ‘solved’ by the Brits in late 1941 and most traffic in the period 1942-45 was solved. Agents abroad relied on hand ciphers, mainly substitution systems. Again most of these could be solved by Bletchley Park during the war.

Through signals intelligence the Brits were able to learn quite a lot about the German spy networks and the Abwehr OOB.
Notice that their operation was targeted, they didn’t intercept everything nor did they have to treat their own population as a security risk. Mail was checked for secret writing and microdots but in this case we are talking about a time of war not peace like today.

All these measures must have saved Britain! If it wasn’t for the ‘defenders of the realm’ then obviously they’d be speaking Deutsche and eating weisswurst today. Or maybe there is another explanation?
Let’s have a look at the official history ‘British Intelligence in the Second World War: Volume 4, Security and Counter-Intelligence’ as it should clear things up. In page 280 it says that wartime successes in counterintelligence depended on a combination of factors, the most important being:

Great Britain being an island, it was possible in war-time to impose strict control of entry which could not be easily evaded. The vulnerable back door via the uncontrollable frontier between Northern Ireland and Eire was protected by the Eire government's vigorous action against the IRA and its determination that Eire should not be used as a base for espionage or sabotage against the United Kingdom. Besides this geographical advantage, in 1939 and throughout the war the United Kingdom had a homogeneous population in which patriotism was still regarded as a cardinal virtue and which, apart from a numerically insignificant minority, was deeply hostile to the Nazi regime. What the Security Executive described as the 'different loyalty' of the leadership and indoctrinated cadres of the CPGB helped Germany only incidentally, and only until she attacked the Soviet Union in June,1941.
So maybe instead of intercepting our internet and phone traffic the Americans can follow these simple guidelines:

1). Make sure their borders are secure and work with Canada and Mexico to ensure this.
2). Have faith in the patriotism of their countrymen to report suspicious activity and deny help to evil spies, terrorists, hackers etc.

But doing something so simple would mean the US intelligence budget would need to be much smaller than 50 plus billion, wouldn’t it?

Thursday, September 19, 2013

Greek radio stations 1940

An interesting file can be found in the British archives, folder HW 40/195 ‘Selected correspondence from the archives of OKW/CHI’.

In 1940 the Lauf and Treuenbrietzen stations that intercepted foreign radio traffic for the German High Command’s decryption department - OKW/Chi, were ordered to give special attention to Greek radio traffic.

The report of 19 January 1940 says:

Because of the geographical dispersion of the Greek state in a large number of islands separated from each other by long distances there has been for some time an internal Greek wireless traffic, above all between the mainland and individual islands (e.g. Crete) and probably between island and island also.

Then a list of Greek radio stations follows.

 

Monday, September 16, 2013

WWII Myths – Multitude of German AFV types

Something that is mentioned often online and in popular history books is that the Germans built too many different types of armored vehicles during WWII. If only they had concentrated production on a handful of types they would have produced more AFV’s than they historically did.

For example ‘Why the Allies won’ by Richard Overy says in page 201
‘At one point in the war there were no fewer than 425 different aircraft models and variants in production. By the middle of the war the German army was equipped with 151 different makes of lorry and 150 different motor-cycles. With such a variety it was difficult to produce in mass.’

I have already covered aircraft production here, so this time let’s take a look at tanks and self propelled guns.
I have posted the German production statistics here, using as a source the book ‘Waffen und Geheimwaffen des deutschen Heeres 1933 – 1945

Using that table and calculating what percentage of total production each vehicle’s production represents we get:


 
At first glance the myth seems to hold true! There are many different ‘types’, each representing a small percentage of production. Even the most well known vehicles do not stand out production wise.

For example the Stug III L/48 is at 16.9%, the Panther at 12.5%, the Pz III L/42 at 4.8%, the Pz IV L/48 at 12.5%.
Is that the end of the discussion? Definitely not! The problem for people who claim that the Germans produced too many types is that they are making a mistake in the way different vehicles are counted.

If two different AFV’s share most of their parts then they may be counted separately but in reality they are the same vehicle.
For example the Stug III that was built in large numbers was simply the same vehicle as the Pz III but with a fixed turret. The same was true for the Pz IV and the Stug IV/Jagdpanzer IV.

If we add up the Pz III, Pz IV, Panther and their fixed turret variants Stug III, Stug/Jagd IV/air-defense versions and JagdPanther then we get 74% of total AFV production. Notice that I haven’t added the Hummel and Nashorn that were built with parts from both the Pz III and the Pz IV, if we add them too then the percentage is 77%.
Panzer 38 production adds another 13.3%. This vehicle was produced throughout the war, first as a tank and then as a self propelled gun not because the Germans were awed by its performance but rather because they had captured the production facilities in Czechoslovakia and couldn’t use them for other purposes.

Even the lowly Pz II was built as a self propelled gun during the war in order to use the existing tank production facilities.
Considering all of the above it is obvious that the Germans built the same basic vehicle types during the war. Instead of introducing new types they simply modified existing types.

In 1940-42 production was centered on the Pz III, Pz IV and Stug III. Since Pz III and Stug III were basically the same vehicle that’s two main types plus the Pz 38.
In 1943-45 the Pz IV and Stug III/IV were joined by the Panther, while the Pz III was no longer built as a tank. So instead of two basic vehicle types we have three ( Pz IV-Stug IV- Jagdpanzer IV, Stug III, Panther) plus the Pz 38 in its SPG variant.

The situation per year is as follows, regarding the top three vehicles by production percentage (counting each vehicle separately):

For 1940



1940

Pz III

47%

Pz 38

17%

Pz IV

13%

Sum

77%

For 1941



1941

Pz III

46%

Pz 38

19%

Stug III

14%

Sum

79%

For 1942



1942

Pz III

43%

Pz IV

16%

Pz II

14%

Sum

73%

 
For 1943



1943

Stug III

26%

Pz IV

25%

Panther

16%

Sum

67%


For 1944



1944

Stug III  

24%

Panther

20%

Pz IV

18%

Sum

62%

 

For 1945



1945

Hetzer

27%

Stug III

25%

Jagd IV

14%

Sum

66%

 
If we count vehicles according to general type (so Pz III=Stug III, Pz IV=Stug IV=Jagd IV, Pz 38= Marder III= Hetzer, etc etc ) we get:

For 1940

1940
Pz III
0.48
Pz 38
0.17
Pz IV
0.13
Sum
0.78

 
For 1941

1941
Pz III
0.60
Pz 38
0.19
Pz IV
0.13
Sum
0.92

 
For 1942

1942
Pz III
0.57
Pz IV
0.16
Pz II
0.14
Sum
0.87

For 1943

1943
Stug III
0.31
Pz IV
0.26
Panther
0.16
Sum
0.73

Note that If we add Nashorn and Hummel then the we get 79%.

For 1944

1944
Pz IV
0.33
Stug III
0.25
Panther
0.22
Sum
0.80

For 1945

1945
Pz 38
0.29
Stug III
0.25
Pz IV
0.24
Sum
0.78

Verdict
It is true that the Germans built several different armored vehicle types during WWII. They did so mainly because they already had the production facilities for some of these types and they couldn’t afford to retool them. Instead obsolete tanks like the Pz III and Pz 38 were produced as self propelled guns.

However even under these circumstances three vehicles the Pz III, Pz IV and Panther (plus their fixed turret variants) accounted for most of the German AFV production, at 77%. Add the Pz 38 and you go to 90%.
The idea that they could have produced more if they concentrated on one type (like the Soviets did with the T-34 or the Americans and their M-4 Sherman) is not correct. Production was limited by the existing facilities and the low priority that was given to AFV production in the German war economy.