Today’s edition of TDI Friday Read is a roundup of posts by TDI President Christopher Lawrence exploring the details of tank combat between German and Soviet forces at the Battle of Kursk in 1943. The prevailing historical interpretation of Kursk is of the Soviets using their material and manpower superiority to blunt and then overwhelm the German offensive. This view is often buttressed by looking at the ratio of the numbers of tanks destroyed in combat. Chris takes a deeper look at the data, the differences in the ways “destroyed” tanks were counted and reported, and the differing philosophies between the German and Soviet armies regarding damaged tank recovery and repair. This yields a much more nuanced perspective on the character of tank combat at Kursk that does not necessarily align with the prevailing historical interpretations. Historians often discount detailed observational data on combat as irrelevant or too difficult to collect and interpret. We at TDI believe that with history, the devil is always in the details.
The Dupuy Institute
Excellence in Historical Research and Analysis
The Dupuy Institute
Excellence in Historical Research and Analysis
Tag Soviet Army
Artillery Effectiveness vs. Armor (Part 5-Summary)
[This series of posts is adapted from the article “Artillery Effectiveness vs. Armor,” by Richard C. Anderson, Jr., originally published in the June 1997 edition of the International TNDM Newsletter.]
Posts in the series
Artillery Effectiveness vs. Armor (Part 1)
Artillery Effectiveness vs. Armor (Part 2-Kursk)
Artillery Effectiveness vs. Armor (Part 3-Normandy)
Artillery Effectiveness vs. Armor (Part 4-Ardennes)
Artillery Effectiveness vs. Armor (Part 5-Summary)
Table IX shows the distribution of cause of loss by type or armor vehicle. From the distribution it might be inferred that better protected armored vehicles may be less vulnerable to artillery attack. Nevertheless, the heavily armored vehicles still suffered a minimum loss of 5.6 percent due to artillery. Unfortunately the sample size for heavy tanks was very small, 18 of 980 cases or only 1.8 percent of the total.
The data are limited at this time to the seven cases.[6] Further research is necessary to expand the data sample so as to permit proper statistical analysis of the effectiveness of artillery versus tanks.
NOTES
[18] Heavy armor includes the KV-1, KV-2, Tiger, and Tiger II.
[19] Medium armor includes the T-34, Grant, Panther, and Panzer IV.
[20] Light armor includes the T-60, T-70. Stuart, armored cars, and armored personnel carriers.
Artillery Effectiveness vs. Armor (Part 4-Ardennes)
[This series of posts is adapted from the article “Artillery Effectiveness vs. Armor,” by Richard C. Anderson, Jr., originally published in the June 1997 edition of the International TNDM Newsletter.]
Posts in the series
Artillery Effectiveness vs. Armor (Part 1)
Artillery Effectiveness vs. Armor (Part 2-Kursk)
Artillery Effectiveness vs. Armor (Part 3-Normandy)
Artillery Effectiveness vs. Armor (Part 4-Ardennes)
Artillery Effectiveness vs. Armor (Part 5-Summary)
NOTES
[14] From ORS Joint Report No. 1. A total of an estimated 300 German armor vehicles were found following the battle.
[15] Data from 38th Infantry After Action Report (including “Sketch showing enemy vehicles destroyed by 38th Inf Regt. and attached units 17-20 Dec. 1944″), from 12th SS PzD strength report dated 8 December 1944, and from strengths indicated on the OKW briefing maps for 17 December (1st [circa 0600 hours], 2d [circa 1200 hours], and 3d [circa 1800 hours] situation), 18 December (1st and 2d situation), 19 December (2d situation), 20 December (3d situation), and 21 December (2d and 3d situation).
[16] Losses include confirmed and probable losses.
[17] Data from Combat Interview “26th Infantry Regiment at Dom Bütgenbach” and from 12th SS PzD, ibid.
Artillery Effectiveness vs. Armor (Part 3-Normandy)
[This series of posts is adapted from the article “Artillery Effectiveness vs. Armor,” by Richard C. Anderson, Jr., originally published in the June 1997 edition of the International TNDM Newsletter.]
Posts in the series
Artillery Effectiveness vs. Armor (Part 1)
Artillery Effectiveness vs. Armor (Part 2-Kursk)
Artillery Effectiveness vs. Armor (Part 3-Normandy)
Artillery Effectiveness vs. Armor (Part 4-Ardennes)
Artillery Effectiveness vs. Armor (Part 5-Summary)
NOTES
[10] From ORS Report No. 17.
[11] Five of the 13 counted as unknown were penetrated by both armor piercing shot and by infantry hollow charge weapons. There was no evidence to indicate which was the original cause of the loss.
[12] From ORS Report No. 17
[13] From ORS Report No. 15. The “Pocket” was the area west of the line Falaise-Argentan and east of the line Vassy-Gets-Domfront in Normandy that was the site in August 1944 of the beginning of the German retreat from France. The German forces were being enveloped from the north and south by Allied ground forces and were under constant, heavy air attack.
Artillery Effectiveness vs. Armor (Part 2-Kursk)
German Army 150mm heavy field howitzer 18 L/29.5 battery. [Panzer DB/Pinterest]
[This series of posts is adapted from the article “Artillery Effectiveness vs. Armor,” by Richard C. Anderson, Jr., originally published in the June 1997 edition of the International TNDM Newsletter.]
Posts in the series
Artillery Effectiveness vs. Armor (Part 1)
Artillery Effectiveness vs. Armor (Part 2-Kursk)
Artillery Effectiveness vs. Armor (Part 3-Normandy)
Artillery Effectiveness vs. Armor (Part 4-Ardennes)
Artillery Effectiveness vs. Armor (Part 5-Summary)
Curiously, at Kursk, in the case where the highest percent loss was recorded, the German forces opposing the Soviet 1st Tank Army—mainly the XLVIII Panzer Corps of the Fourth Panzer Army—were supported by proportionately fewer artillery pieces (approximately 56 guns and rocket launchers per division) than the US 1st Infantry Division at Dom Bütgenbach (the equivalent of approximately 106 guns per division)[4]. Nor does it appear that the German rate of fire at Kursk was significantly higher than that of the American artillery at Dom Bütgenbach. On 20 July at Kursk, the 150mm howitzers of the 11th Panzer Division achieved a peak rate of fire of 87.21 rounds per gum. On 21 December at Dom Bütgenbach, the 155mm howitzers of the 955th Field Artillery Battalion achieved a peak rate of fire of 171.17 rounds per gun.[5]
NOTES
[4] The US artillery at Dom Bütgenbach peaked on 21 December 1944 when a total of 210 divisional and corps pieces fired over 10,000 rounds in support of the 1st Division’s 26th Infantry.
[5] Data collected on German rates of fire are fragmentary, but appear to be similar to that of the American Army in World War ll. An article on artillery rates of fire that explores the data in more detail will be forthcoming in a future issue of this newsletter. [NOTE: This article was not completed or published.]
Notes to Table I.
[8] The data were found in reports of the 1st Tank Army (Fond 299, Opis‘ 3070, Delo 226). Obvious math errors in the original document have been corrected (the total lost column did not always agree with the totals by cause). The total participated column evidently reflected the starting strength of the unit, plus replacement vehicles. “Burned'” in Soviet wartime documents usually indicated a total loss, however it appears that in this case “burned” denoted vehicles totally lost due to direct fire antitank weapons. “Breakdown” apparently included both mechanical breakdown and repairable combat damage.
[9] Note that the brigade report (Fond 3304, Opis‘ 1, Delo 24) contradicts the army report. The brigade reported that a total of 28 T-34s were lost (9 to aircraft and 19 to “artillery”) and one T-60 was destroyed by a mine. However, this report was made on 11 July, during the battle, and may not have been as precise as the later report recorded by 1st Tank Army. Furthermore, it is not as clear in the brigade report that “artillery” referred only to indirect fire HE and not simply lo both direct and indirect fire guns.
Artillery Effectiveness vs. Armor (Part 1)
[This series of posts is adapted from the article “Artillery Effectiveness vs. Armor,” by Richard C. Anderson, Jr., originally published in the June 1997 edition of the International TNDM Newsletter.]
Posts in the series
Artillery Effectiveness vs. Armor (Part 1)
Artillery Effectiveness vs. Armor (Part 2-Kursk)
Artillery Effectiveness vs. Armor (Part 3-Normandy)
Artillery Effectiveness vs. Armor (Part 4-Ardennes)
Artillery Effectiveness vs. Armor (Part 5-Summary)
The effectiveness of artillery against exposed personnel and other “soft” targets has long been accepted. Fragments and blast are deadly to those unfortunate enough to not be under cover. What has also long been accepted is the relative—if not total—immunity of armored vehicles when exposed to shell fire. In a recent memorandum, the United States Army Armor School disputed the results of tests of artillery versus tanks by stating, “…the Armor School nonconcurred with the Artillery School regarding the suppressive effects of artillery…the M-1 main battle tank cannot be destroyed by artillery…”
This statement may in fact be true,[1] if the advancement of armored vehicle design has greatly exceeded the advancement of artillery weapon design in the last fifty years. [Original emphasis] However, if the statement is not true, then recent research by TDI[2] into the effectiveness of artillery shell fire versus tanks in World War II may be illuminating.
The TDI search found that an average of 12.8 percent of tank and other armored vehicle losses[3] were due to artillery fire in seven eases in World War II where the cause of loss could be reliably identified. The highest percent loss due to artillery was found to be 14.8 percent in the case of the Soviet 1st Tank Army at Kursk (Table II). The lowest percent loss due to artillery was found to be 5.9 percent in the case of Dom Bütgenbach (Table VIII).
The seven cases are split almost evenly between those that show armor losses to a defender and those that show losses to an attacker. The first four cases (Kursk, Normandy l. Normandy ll, and the “Pocket“) are engagements in which the side for which armor losses were recorded was on the defensive. The last three cases (Ardennes, Krinkelt. and Dom Bütgenbach) are engagements in which the side for which armor losses were recorded was on the offensive.
Four of the seven eases (Normandy I, Normandy ll, the “Pocket,” and Ardennes) represent data collected by operations research personnel utilizing rigid criteria for the identification of the cause of loss. Specific causes of loss were only given when the primary destructive agent could be clearly identified. The other three cases (Kursk, Krinkelt, and Dom Bütgenbach) are based upon combat reports that—of necessity—represent less precise data collection efforts.
However, the similarity in results remains striking. The largest identifiable cause of tank loss found in the data was, predictably, high-velocity armor piercing (AP) antitank rounds. AP rounds were found to be the cause of 68.7 percent of all losses. Artillery was second, responsible for 12.8 percent of all losses. Air attack as a cause was third, accounting for 7.4 percent of the total lost. Unknown causes, which included losses due to hits from multiple weapon types as well as unidentified weapons, inflicted 6.3% of the losses and ranked fourth. Other causes, which included infantry antitank weapons and mines, were responsible for 4.8% of the losses and ranked fifth.
NOTES
[1] The statement may be true, although it has an “unsinkable Titanic,” ring to it. It is much more likely that this statement is a hypothesis, rather than a truism.
[2] As pan of this article a survey of the Research Analysis Corporation’s publications list was made in an attempt to locate data from previous operations research on the subject. A single reference to the study of tank losses was found. Group 1 Alvin D. Coox and L. Van Loan Naisawald, Survey of Allied Tank Casualties in World War II, CONFIDENTIAL ORO Report T-117, 1 March 1951.
[3] The percentage loss by cause excludes vehicles lost due to mechanical breakdown or abandonment. lf these were included, they would account for 29.2 percent of the total lost. However, 271 of the 404 (67.1%) abandoned were lost in just two of the cases. These two cases (Normandy ll and the Falaise Pocket) cover the period in the Normandy Campaign when the Allies broke through the German defenses and began the pursuit across France.
TDI Friday Read: Measuring The Effects of Combat in Cities
Between 2001 and 2004, TDI undertook a series of studies on the effects of urban combat in cities for the U.S. Army Center for Army Analysis (CAA). These studies examined a total of 304 cases of urban combat at the divisional and battalion level that occurred between 1942 and 2003, as well as 319 cases of concurrent non-urban combat for comparison.
The primary findings of Phases I-III of the study were:
- Urban terrain had no significantly measurable influence on the outcome of battle.
- Attacker casualties in the urban engagements were less than in the non-urban engagements and the casualty exchange ratio favored the attacker as well.
- One of the primary effects of urban terrain is that it slowed opposed advance rates. The average advance rate in urban combat was one-half to one-third that of non-urban combat.
- There is little evidence that combat operations in urban terrain resulted in a higher linear density of troops.
- Armor losses in urban terrain were the same as, or lower than armor losses in non-urban terrain. In some cases it appears that armor losses were significantly lower in urban than non-urban terrain.
- Urban terrain did not significantly influence the force ratio required to achieve success or effectively conduct combat operations.
- Overall, it appears that urban terrain was no more stressful a combat environment during actual combat operations than was non-urban terrain.
- Overall, the expenditure of ammunition in urban operations was not greater than that in non-urban operations. There is no evidence that the expenditure of other consumable items (rations; water; or fuel, oil, or lubricants) was significantly different in urban as opposed to non-urban combat.
- Since it was found that advance rates in urban combat were significantly reduced, then it is obvious that these two effects (advance rates and time) were interrelated. It does appear that the primary impact of urban combat was to slow the tempo of operations.
In order to broaden and deepen understanding of the effects of urban combat, TDI proposed several follow-up studies. To date, none of these have been funded:
- Conduct a detailed study of the Battle of Stalingrad. Stalingrad may also represent one of the most intense examples of urban combat, so may provide some clues to the causes of the urban outliers.
- Conduct a detailed study of battalion/brigade-level urban combat. This would begin with an analysis of battalion-level actions from the first two phases of this study (European Theater of Operations and Eastern Front), added to the battalion-level actions completed in this third phase of the study. Additional battalion-level engagements would be added as needed.
- Conduct a detailed study of the outliers in an attempt to discover the causes for the atypical nature of these urban battles.
- Conduct a detailed study of urban warfare in an unconventional warfare setting.
Details of the Phase I-III study reports and conclusions can be found below:
Measuring the Effects of Combat in Cities, Phase II – part 1
Measuring the Effects of Combat in Cities, Phase II – part 2
Measuring the Effects of Combat in Cities, Phase III – part 1
Measuring the Effects of Combat in Cities, Phase III – part 2
Measuring the Effects of Combat in Cities, Phase III – part 2.1
Measuring the Effects of Combat in Cities, Phase III – part 3
Measuring the Effects of Combat in Cities, Phase II – part 2
There was actually supposed to be a part 2 to this Phase II contract, which was analysis of urban combat at the army-level based upon 50 operations, of which a half-dozen would include significant urban terrain. This effort was not funded.
On the other hand, the quantitative analysis of battles of Kharkov only took up the first 41 pages of the report. A significant part of the rest of the report was a more detailed analysis and case study of the three fights over Kharkov in February, March and August of 1943. Kharkov was a large city, according to the January 1939 census, it has a population of 1,344,200, although a Soviet-era encyclopedia gives the pre-war population as 840,000. We never were able to figure out why there was a discrepancy. The whole area was populated with many villages. The January 1939 gives Kharkov Oblast (region) a population of 1,209,496. This is in addition to the city, so the region had a total population of 2,552,686. Soviet-era sources state that when the city was liberated in August 1943, the remaining population was only 190,000. Kharkov was a much larger city than any of the others ones covered in Phase I effort (except for Paris, but the liberation of that city was hardly a major urban battle).
The report then does a day-by-day review of the urban fighting in Kharkov. Doing a book or two on the battles of Kharkov is on my short list of books to write, as I have already done a lot of the research. We do have daily logistical expenditures of the SS Panzer Corps for February and March (tons of ammo fired, gasoline used and diesel used). In March when the SS Panzer Corps re-took Kharkov, we noted that the daily average for the four days of urban combat from 12 to 15 March was 97.25 tons of ammunition, 92 cubic meters of gasoline and 10 cubic meters of diesel. For the previous five days (7-11 March) the daily average was 93.20 tons of ammunition, 145 cubic meters of gasoline and 9 cubic meters of diesel. Thus it does not produce a lot of support for the idea that–as has sometimes been expressed (for example in RAND’s earlier reports on the subject)–that ammunition and other supplies will be consumed at a higher rate in urban operations.
We do observe from the three battles of Kharkov that (page 95):
There is no question that the most important lesson found in the three battles of Kharkov is that one should just bypass cities rather than attack them. The Phase I study also points out that the attacker is usually aware that faster progress can be made outside the urban terrain, and that the tendency is to weight one or both flanks and not bother to attack the city until it is enveloped. This is indeed what happened in two of the three cases at Kharkov and was also the order given by the Fourth Panzer Army that was violated by the SS Panzer Corps in March.
One must also note that since this study began the United States invaded Iraq and conducted operations in some major urban areas, albeit against somewhat desultory and ineffective opposition. In the southern part of Iraq the two major port cities Umm Qasar and Basra were first enveloped before any forces were sent in to clear them. In the case of Baghdad, it could have been enveloped if sufficient forces were available. As it was, it was not seriously defended. The recent operations in Iraq again confirmed that observations made in the two phases of this study.
P.S. The picture is of Kharkov in 1942, when it was under German occupation.
Measuring the Effects of Combat in Cities, Phase II – part 1
Our first urban warfare report that we did had a big impact. It clearly showed that the intensity of urban warfare was not what some of the “experts” out there were claiming. In particular, it called into question some of the claims being made by RAND. But, the report was based upon Aachen, Cherbourg, and a collection of mop-up operations along the Channel Coast. Although this was a good starting point because of the ease of research and availability of data, we did not feel that this was a fully representative collection of cases. We also did not feel that it was based upon enough cases, although we had already assembled more cases than most “experts” were using. We therefore convinced CAA (Center for Army Analysis) to fund a similar effort for the Eastern Front in World War II.
For this second phase, we again assembled a collection of Eastern Front urban warfare engagements in our DLEDB (Division-level Engagement Data Base) and compared it to Eastern Front non-urban engagements. We had, of course, a considerable collection of non-urban engagements already assembled from the Battle of Kursk in July 1943. We therefore needed a good urban engagement nearby. Kharkov is the nearest major city to where these non-urban engagements occurred and it was fought over three times in 1943. It was taken by the Red Army in February, it was retaken by the German Army in March, and it was taken again by the Red Army in August. Many of the units involved were the same units involved in the Battle of Kursk. This was a good close match. It has the additional advantage that both sides were at times on the offense.
Furthermore, Kharkov was a big city. At the time it was the fourth biggest city in the Soviet Union, being bigger than Stalingrad (as measured by pre-war population). A picture of its Red Square in March 1943, after the Germans retook it, is above.
We did have good German records for 1943 and we were able to get access to Soviet division-level records from February, March and August from the Soviet military archives in Podolsk. Therefore, we were able to assembled all the engagements based upon the unit records of both sides. No secondary sources were used, and those that were available were incomplete, usually one-sided, sometimes biased and often riddled with factual errors.
So, we ended up with 51 urban and conurban engagements from the fighting around Kharkov, along with 65 non-urban engagements from Kursk (we have more now).
The Phase II effort was completed on 30 June 2003. The conclusions of Phase II (pages 40-41) were similar to Phase I:
.Phase II Conclusions:
- Mission Accomplishment: This [Phase I] conclusion was further supported. The data does show a tendency for urban engagements not to generate penetrations.
- Casualty Rates: This [Phase I] conclusion was further supported. If urban combat influenced the casualty rate, it appears that it resulted in a reduction of the attacker casualty rate and a more favorable casualty exchange ratio compared to nonurban warfare. There still appears to be no basis to the claim that urban combat is significantly more intense with regards to casualties than is nonurban warfare.
- Advance Rates: There is no strong evidence of a reduction in the advance rates in urban terrain in the Eastern Front data. TDI still stands by its original conclusion that the average advance rate in urban combat should be one-half to one-third that of nonurban combat.
- Linear Density: Again, there is little evidence that the presence of urban terrain results in a higher linear density of troops, but unlike the ETO data, the data did not show a tendency to trend in that direction.
- Armor Losses: This conclusion was further supported (Phase I conclusion was: Overall, it appears that the loss of armor in urban terrain is the same as or less than that found in nonurban terrain, and in some cases is significantly lower.)
- Force Ratios: The conclusion was further supported (Phase I conclusion was: Urban combat did not significantly influence the Force Ratio required to achieve success or effectively conduct combat operations).
- Duration of Combat: Nothing could be determined from an analysis of the data regarding the Duration of Combat (Time) in urban versus nonurban terrain.
There is a part 2 to this effort that I will pick up in a later post.
Why Do Americans Hate Military Theory?
In his new book, On Tactics: A Theory of Victory in Battle, Brett Friedman wrote:
[The] lack of strategic education has produced a United States military adrift. A cottage industry of shallow military thought attached itself to the Department of Defense like a parasite, selling “new” concepts that ranged from the specious (such as the RMA and effects-based operations), to the banal (like “hybrid” and “asymmetric” warfare), to the nonsensical (like 4th Generation Warfare and Gray Zone/Wars). An American officer corps, bereft of a solid understanding of strategic theory, seizes on concept after concept, seeking the next shiny silver bullet that it can fire to kill the specter of strategic disarray.
The U.S. military establishment’s general disregard and disinterest in theorizing about war and warfare is not new. Trevor Dupuy was also critical of the American approach to thinking about theory, especially its superficial appreciation for the value of military history. As he wrote in Understanding War: History and Theory of Combat (1987):
In general, and with only a few significant exceptions, until very recently American military theorists have shown little interest in the concept of a comprehensive theory or science of combat. While most Americans who think about such things are strong believers in the application of science to war, they seem not to believe, paradoxically, that waging war can be scientific, but that it is an art rather than a science. Even scientists concerned with and involved in military affairs, who perhaps overemphasize the role of science in war, also tend to believe that war is a random process conducted by unpredictable human beings, and thus not capable of being fitted into a scientific theoretical structure. [p. 51]
Like Friedman, Dupuy placed a good deal of the blame for this on the way U.S. military officers are instructed. He saw a distinct difference in the approach taken in the U.S. versus the way it was used by the (then) Soviet Union. In a 1989 conference paper, he contended that:
The United States Armed Forces pay lip service to the importance of military history. Officers are urged to read military history, but given little guidance on how military history can be really useful to them. The fundamental difference between the Soviet approach and the American approach, as I see it, is that the American officer is invited (but not really encouraged) to be a military history dilettante. The Soviets seriously study, and use military history. Figure 1 summarizes the differences in approaches of the U.S. and the Soviet armed forces to military history analysis.
Dupuy devoted an entire chapter of Understanding War to the Soviet scientific approach to the study and application of warfare. There was a time when the mention of Soviet/Russian military theory would have produced patronizing smirks from American commentators. In truth, Russian military theorizing has a long and robust tradition; much deeper than its American counterpart. Given the recent success Russia has had in leveraging its national security capabilities to influence favorable geopolitical outcomes, it might be that those theories are useful after all. One need not subscribe to the Soviet scientific approach to warfare to acknowledge the value of a scientific approach to studying warfare.