Arctic Dragon: M47 Dragon trials in Norway, 1982

An introduction to the M47 in Korps Mariniers service

To recapitulate the previous articles here, here, and here, the M47 Dragon was the Dutch Army’s medium range anti-tank weapon, occupying the following position in the “anti tank mix”:

Purpose System Launch unit cost, guilders (year) Ammunition cost, guilders (year)
Self-defence M72 LAW 215 (1972) / 600 (1981)
Short range M2 Carl Gustaf 3.250 (1963) 400 (1978)
Medium range M47 Dragon 145.000 (1976) 9.000 (1976)
Long range TOW 1035.000 (1976) 17.500 (1976)

The trials presented below, however, were undertaken by the Korps Mariniers, who operated under different conditions from their Army counterparts. Earmarked for NATO’s Northern flank in Norway, it was noted how “the fields of fire in the operational theatre are greater than 1000 metres.” This is to be contrasted with the Army, which noted how: “In the operational theatre of 1 (NL) Corps, 70% of terrain is limited to fields of fire below 1000 metres”.

It was in this context that the Dragon would replace the 106mm M40 recoilless rifle, a weapon found in the battalion’s weapon company, equipping a single anti-tank platoon.

The Amphibious Combat Group (AGGP) possessed a single anti-tank platoon (atpel) in its weapons company (ostcie)

The anti-armour weapons of both the Army and Korps Mariniers are summarised as such:

Type Max. Eff. Range Army scale of issue Korps Mariniers scale of issue
TOW 3.000 m 6×, later 12×/16× per battalion None
Tanks 2.800 m 104× per armoured brigade
61× per armoured infantry brigade
106mm M40 1.200 m 6× per battalion,
replaced by TOW
6× per battalion,
replaced by M47 Dragon
M47 Dragon 1.000 m 2× per platoon 21× per battalion
84mm Carl Gustav 500 m 2× per platoon,
supplemented by M47 Dragon and pushed down to 1× squad
1× platoon
66mm LAW 200 m Issued as ammunition, 253× per armoured infantry battalion Issued as ammunition

Original document, translated

To: the commander of the Korps Mariniers
You are hereby presented with a report concerning trials of the M47 Dragon weapon system under arctic conditions during Clockwork ’82.

1. Purpose of the trials:

a. To investigate the weapon system’s behaviour in arctic environments;
b. To investigate possible interference to the weapon system;
c. To investigate the influence of interferences on the guidability of the wire-guided rocket;
d. To gather information which concerns the tactical use of the weapon system.

2. Set-up of the trials:

A trial schedule was created to investigate a number of phenomena. This was done under informal cooperation with the Materieelsbeproevingsafdeling – 1, located in Zaandam.

Round number 1-6 7-11 12-16 17-21 22-27 28-32 33-37 38-42 Characteristic trialled
x x x x - - - - low temperature
- - - - x x x x relatively high-low temperatures
1x 1x 1x 1x 2x 2x 2x 2x number of thermal shocks
x - x x - x drop tests
x x x x transported in BV 202
6 1, 2 7, 8 12, 13 17, 18 Day 1
10 3, 4, 5 19, 20 22, 23, 24 38, 39 Day 2
8 9, 10 14, 15 28, 29 33, 34 Day 3
8 25, 26 30, 31 35, 36 40, 41 Day 4
8 6 11 16 21 27 32 37 42 Day 5
42 6 5 5 5 6 5 5 5 Total

3. Execution of the trials:

a. Rounds were fired by gunners who had qualified “0068” or “0069” in the past 3 to 8 months, all of whom had already fired the M223 rocket at the ISK Harskamp.
b. All rounds were fired at a stationary target from lanes 13 and 14 at the Drevja range, Norway. The target consisted of a wooden frame covered in burlap, measuring 3 metres wide and 1.8 metres high. Firing took place at distances of 700 (lane 14) and 800 (lane 13) metres respectively.
c. Before firing the weapon system, the launch units were laid out in the snow for several hours. The trackers were continuously stored in unheated spaces for the duration of the trial, and were inspected both during and after trials using a tracker sight testing unit.
d. Every shot fired was captured by a technical photographer of the Marine Electronisch en Optisch Bedrijf, Wassenaar.
e. Every shot fired was evaluated by means of an interview during which all observations and results were recorded in a trial report. An example of the used form is included in the appendix.
f. Of the 42 launch units trialled, only 26 rounds were fired during the final phase of trials as a result of relatively high outdoor temperatures (±0). This creates a certain bias with regards to the reliability of conclusions that can be made based on the firing results and observations.
g. The 42 packed launcher units were stored in an unheated bunker in Drevja for a duration of 3 to 4½ weeks. Upon inspection of the launch units, the moisture indicators on the forward shock absorber were discoloured to “ROSE 30”, meaning the weapon can be fired but must be stored in a different manner (see VS 7-611 Koninklijke Landmacht and Zakboek voor de schutter Korps Mariniers p. 14). As a result of continuously changing weather and temperature conditions during the trial period, the number of thermal shocks (transitions warm-cold/cold-warm) could not be accurately determined.

4. Trial results:

a. No interference or defects to the used tracker sights were noted. Breathing on the ocular must however be avoided.
b. Of the 26 rounds fired, 14 would hit compared to 12 misses. An overview of hits/misses and meteorological circumstances can be found on the following page.

Day Avg. Temp. Precipitation Wind Hit Miss
080282 - 14°C None None 4 4
090282 - 4°C None None 7 3
170282 + 1°C None None 3 5

c. During trials, the following phenomena were noted:

1. Excessive flash

Excessive flash was noted sixteen times upon the activation of the propelling charge and after the launching of the rocket from the launch tube. The following processes took place inside the launch tube (for a maximum of 2 seconds):
- Activation of the propelling charge and expulsion of the rocket
- Vacuum phase
- The formation of flashes from the rear to the front (“leakage”).
The above-mentioned “leakage” flashes form as a result of the propelling charge’s powder particles afterburning inside the launch tube, and last for 1 to 2 seconds. The target’s visibility is influenced by these flashes, causing the gunner to see a yellow haze through his sight. It could not be established as to whether the “leakage” flashes influenced the infra-red mechanism present on the sight.

2. Excessive smoke

Due to the rearward flow of the propellant charge’s gasses, a plume of smoke has been noted behind the firing position 22 times. This white-grey plume of smoke was visible between 3 to 24 seconds.

3. Uncontrolled flight

Seven times, the rocket’s flight was uncontrolled. During one occurrence, the rocket thrusters were not activated after activation of the propelling charge and expulsion of the rocket from the launch tube. Upon further inspection of the guidance wire and the rocket, it was found the guidance wire was intact. The 60 rocket thrusters were still sealed.
In another single instance, the guidance wire broke (see point 4).
During one occurrence, it was suspected the gyroscopic mechanism had jammed. The rocket followed an upward trajectory, during which the rocket thrusters only gave off upward corrections. The rocket travelled a distance of 1400m. Such a flight path has been previously observed during trials by MBA-1.
Four instances have been noted where the rocket barely or did not respond to the gunner’s corrections.

4. Broken guidance wire

In one instance, the guidance wire broke off. The wire had broken inside the launch tube (30 cm of wire remaining), with the other section of wire (1050 metres) still attached to the rocket.

5. Wire pulling

Nine instances were noted where the rocket pulled on the launch tube during its flight. Instead of the wire unspooling in an even manner, the rocket would impose pulling and shocking motions upon the launch tube after having travelled 600 metres.

6. Delayed response from the rocket

It was noted five times how the rocket would respond to corrections with a delay. The rocket would (at both short and long range) swerve to the left and right of the line of sight. This caused the flight path and the line of sight to misalign during the final part of the flight path, causing three misses. Two hits were however noted, presumably due to the target and the rocket’s swerving motion aligning. The delayed response was only noted during the first day of firing (average temperature, - 14 °C).

5. Conclusions

a. During trials of the M47 Dragon weapon system, rocket type M223 it was shown the weapon behaves differently from trials in the Netherlands. Phenomena not previously noted were: excessive smoke and flash, wire pulling, delayed responses of the rocket.
b. Launches and flight path were not without issue. A number of the issues have also been noted in the Netherlands: broken guidance wires and jammed gyro mechanisms.
c. The influence of the pre-trial treatment of the Dragon does not appear to significantly influence hit probability. The number of thermal shocks could not be determined, the height at which the launch units were dropped was 75 cm (surface: frozen snow), the launch units were horizontally stored in the BV 202 in their racks.
d. The negative effect of being stored in moist conditions can not be ruled out with regards to the launch units’ functioning.
e. The issues and noted phenomena during the launch and flight of the rocket influence the system’s hit probability. The hit probability at temperatures below freezing (down to – 32 °C) could be lower than previously suspected.
f. Phenomena such as excessive smoke and flash could be of influence for the tactical use of Dragon-equipped anti-tank units. The firing position may possibly be easily recognized.

6. Recommendations

a. In light of the experience gathered with the M47 Dragon M223 in Norway, the conditions under which Dragon launch units M222 (especially war stocks) and M223 are stored in the Netherlands must be examined.
b. As the trials were incomplete, it is recommended further trials are undertaken during Clockwork 1983 in Norway, testing the Dragon’s behaviour at temperatures between -15 °C and - 32 °C.

Annex 1

1. Round number : 8. Precipitation : None/fog/rain/snow
2. Date : 9. Visibility : Good/mediocre/poor
3. Time : 10. Range to target :                                                             m
4. Lane : 11. Direction :                                                             °
5. Temperature :                                                             °C 12. Stationary/moving target :
6. Wind speed :                                                             m/s 13. Gunner :
7. Wind direction :                                                             ° 14. Firing position : sitting/kneeling/standing

15. Hit/miss (reason) :
16. Misfire (reason) :
17. Gunner comments through the sight

a. After firing

(1) Formation of fog: yes/no
(2) Target visibility: good/mediocre/poor
(3) Target visible after: ... sec
(4) Target visible after: ... metres
(5) Notes:

b. During the rocket’s flight:

(1) Formation of fog: yes/no
(2) Target visibility: good/mediocre/poor
(3) Rocket’s controllability: good/mediocre/uncontrollable
(4) Notes:

18. Assistant gunner 1/2 comments

a. Position : ... metres to the left/right of the firing position
b. Formation of fog in proximity of the firing position: none/all around/sides/front/rear
c. Duration of fog: ... sec
d. Snow cloud in proximity of the firing position: none/all around/sides/front/rear
e. Duration of snow cloud: ... sec
f. Further comments:

19. Assistant gunner 3 comments

a. Position : ... metres to the left/right of the firing position
b. Formation of fog due to thrusters: yes/no
c. Duration of fog: ... sec
d. Firing position identifiable as a result of fog: yes/no
e. Further comments:

20. Deviation from trial schedules:
21: Further comments:

The Clockwork '82 Trials: summary and conclusion

Round number Treatment: Firing: Peculiarities: Technical   Peculiarities:
Once in snow Twice in snow Dropped Transported Dropped + transported Moisture indicators ROSE 30 Temperature °C Hit Miss Miss caused by: Excessive Flash Excessive Smoke Uncontrolled Flight Broken Guidance Wire Wire Pulling Delayed response Rocket
Suspected Guidance
Error Gunner
Ammunition Ammunition, suspected Undeterminable
Day 1 1 x - - - - x -15 - x - x - - - x x x - -
2 x - - - - x -14 x - - - - - x x - - x -
7 x - x - - x -14 - x - - - x - x - - x x
8 x - x - - x -14 - x - - x - - x x - - x
12 x - - x - x -15 - x - - - x x x - - - x
13 x - - x - x -16 x - - - - - - x - - - -
17 x - - - x x -14 x - - - - - - x - - - x
18 x - - - x x -15 x - - - - - x x - - x x
Day 2 3 x - - - - x -3 - x - x - - x x x - - -
4 x - - - - x -3 x - - - - - x x - - x -
5 x - - - - x -3 x - - - - - x x - - - -
19 x - - - x x -4 x - - - - - x - - - - -
20 x - - - x x -4 x - - - - - - - - - - -
22 - x - - - x -4 x - - - - - x x - - - -
23 - x - - - x -4 - x x - - - x x - - - -
24 - x - - - x -4 x - - - - - x x - - - -
38 - x - - x x -4 - x x - - - x - - - - -
39 - x - - x x -4 x - - - - - x - - - - -
Day 3 9 x - x - - x 0 - x - - x - - x x - - -
10 x - x - - x 0 - x - - - x x x - - x -
14 x - - x - x +1 - x - - x - - x x - x -
15 x - - x - x +1 - x - - x - - x x - x -
28 - x x - - x +1 x - - - - - x x - - x -
29 - x x - - x +1 x - - - - - - x - - x -
33 - x - x - x 0 x - - - - - x x - - - -
34 - x - x - x 0 - x - x - - x x x - - -
Sum: 17 9 6 6 6 26 14 12 2 3 4 3 16 22 7 1 9 5

The trials’ findings are, ultimately, hard to analyse as a result of the small sample size of only 26 rounds fired.

First, a note on hit rates under normal conditions: During trials against the Milan, a hit rate of 89 percent was achieved (compared to 76 percent on the Milan) firing an unknown number of rounds. During further Army testing, a hit rate of 91.67 percent was achieved with 24 rounds fired. Finally, the Korps Mariniers noted a general hit rate of 85 percent.

With the above in mind, we can evaluate the trial results and examine the effect of improper storage, exposure to snow and weather, and ammunition being drop-tested.

Out of 26 rounds fired, 14 hit their target, with 12 misses due to a number of observed, suspected, and undeterminable causes. It is these misses with a suspected and undeterminable cause which further limit the conclusions that can be drawn from the data set.

First, we will assume that the suspected guidance errors on the part of the gunner were indeed caused by the gunner.

It must be noted that operator error is a problematic factor to assess. The Dragon has been noted as a “difficult weapon to fire, and firing trials took place under “optimal, completely schoolish peacetime conditions, with safety considerations associated with firing ranges forming the most limiting factor. As such, operator error could very well be an issue with the operator, training, or weapon itself. The Korps Mariniers was very much aware of a potentially decreased hit rate under combat conditions, and was for that matter interested in acquiring the M175 mount, suggesting improvements were welcome in spite of high peacetime hit rates.

Next, there are seven instances of a miss caused by a (suspected) ammunition failure.

As the problem of broken guidance wires had been noted during earlier trials, we can somewhat safely state this issue was inherent to the Dragon and not necessarily related to the nature of the trials. The same applies to the jammed gyroscope, leaving us with five instances of uncontrolled flight: one failure of the thrusters to activate, and four instances of little or no response to gunner input. Neither issue had been present during previous trials.

It must be noted that the document does not note in which cases the above seven instances of uncontrolled flight overlap with the issue of wire pulling or a delayed response.

Finally, there are three instances in which an undeterminable cause led to a miss. These rounds fired suffered from wire pulling, a delayed response, and both wire pulling and a delayed response combined respectively. It is assumed these issues are caused by the nature of the trial: both issues were entirely new.

It is worth considering the impact of both issues on the hit rate separately: out of 9 rounds fired suffering from wire pulling, 5 would still find their way to the target, whilst the same is true for 2 out of 5 rounds suffering from a delayed response to the gunner’s input.

Hit/miss Cause Technical peculiarities
Total hits 14
Hits despite a delayed response 1
Hits despite wire pulling 4
Hits despite both wire pulling and a delayed response 1
Misses 12
Suspected guidance error gunner 2
Uncontrolled flight 7
Broken guidance wire 1
Jammed gyro 1
No activation of thrusters 1
Little/no response to gunner input 4
Undeterminable cause, assumed ammunition issues 3
Wire pulling 1
Delayed response to gunner input 1
Wire pulling combined with a delayed response to gunner input 1
Total rounds fired 26

With 22 out of 26 rounds suffering from an excessive amount of smoke upon launch, it can almost certainly be stated the inappropriate storage conditions relate to this, especially as the issue had not been noted during trials in the Netherlands under normal conditions. The same applies to the matter of excessive flash – affecting 16 out of 22 rounds fired. The issue of wires snapping, something commented on as occurring frequently, does not at all appear to have been as great of a problem, occurring once during firing.

The conclusion, especially with regards to incorrect storage in unheated, moist conditions, is appropriately vague: “The negative effect of being stored in moist conditions can not be ruled out with regards to the launch units’ functioning”.

With all of the above in mind, it can be said with certainty that the call for further testing the following year is appropriate.

Speaking more generally on the Korps Mariniers' use of the M47, the Dragon was a particularly awkward weapon as a battalion-level anti-tank system, as it was simply never intended to be used as such. Indeed, it comes across as slightly comical for a fully motorised anti-tank platoon to use a system fully intended to be man-portable, and optimised for light weight.

Whilst perfectly satisfying the Army's 1000m requirement for a platoon-level system (indeed, later pushed down to the squad and fireteam-level), it was not suited to the Mariniers' operational theatre of Norway.

The issues noted in the arctic environment further reinforce this point, as these were of no concern to the Army.
In this case, it would seem joint procurement with the Dutch Army limited the Korps Mariniers to a great extent.
Joint procurement with the British – who operated the Milan – would seem more logical and was indeed considered, although the Milan came with certain logistical and financial issues which have been explored in previous articles.

Original document