EFFECTS OF SPHERICAL FRONT WINDOW ON DIVER PERFORMANCE

K Strannemalm1, A Hägglin1, and H Örnhagen2
1 Dept of Naval Architecture and Ocean Engineering, Chalmers University of Technology, Göteborg 2 National Defence Research Establishment, Naval Med Div, Horsfjärden Sweden

EUBS 90, Amsterdam

ABSTRACT
Today, a saturation diver is obstructed by stiff hoses for gas reclaim, breathing line insulation and bail-out system on his helmet and can therefore not turn his head. This together with a small front window, makes it difficult for him to survey the surroundings. A solution to this problem is a spherical acrylic helmet, but rumors among divers have caused doubts regarding diver performance, using this type of helmet and only comparatively few helmets of this type are in use.

To evaluate effects from optical distortion in a spherical acrylic helmet, six tests were desinged, the objective being to discuss problems in distance and angular determination. A questionnaire on subjective opinions was also given to the test subjects.

No difference in performance was seen between flat and spherical visors. In general, distances were underestimated both in flat and spherical visors. Only few negative comments were made on the spherical visor helmet, most of them related to the high boyancy of the prototype test helmet. We conclude, everything else being equal, that there is no obvious performance limitaiton using spherical helmets.

INTRODUCTION
As a part of the development of a new breathing apparatus for ultra deep diving, the SC 450 [2], tests have been performed to establish the effects on diver performance by optical distortion, depending on the use of a double curved visor on a diving helmet. The tests were carried out at the Department of Underwater Technology at Chalmers University of Technology in cooperation with Advanced Underwater Technology AB, owner of the patents concerning SC 450.

In professional diving, divers visual fields are often restricted. Small front windows in hard hats were compensated for by side and top windows. Modern helmets and full face masks with only one face port, allowed in the original design, movement of the head to improve the field of vision. However, today´s requirements for bail-out, gas reclaim, and gas heating have more or less eliminated head movements. Restriction in vision is not only a safety risk, but it also effects the efficiency of the diver.

A possible solution to this problem is a spherical acrylic helmet in which the diver can move his head. The technology concerning this type of materials has reached far when it comes to impact strenght and optical performance. If an acrylic surface is scratched does not matter, since most marks will be invisible under water. If a rebreathing circuit is used, problems will occur with condensation of water on the inside of the sphere. Earlier this has been solved with a "windscreen" wiper on the inside, manouvered by the diver [3]. Today, special coatings that prevent misting are available. Due to the shape of the sphere, the buoayant force will be evident. It can, however, be kept at a minimum through careful design.

Some spheric helmets have been sold, but rumors among divers have in many cases caused doubts regarding divers performance, using this type of helmets. The rumors are more widely spread than the sphereical helmets and thus it is impossible for most divers, with a view in this matter, to have tried such a helmet. A possible explanation is what we call "the inverted goldfish bowl hypothesis". The distorted sight of a goldfish can easily make you believe that everything on the outside of the bowl must look twisted, viewed from the inside. The view from the inside of a gasfilled "goldfish bowl" into water is, however, not as disturbed as one might believe.

The French LBS system uses a spherical transparent polycarbonate helmet. The main user of the system, Oceaneering, have however, removed the breathing circuit from the bubble and placed it inside their conventional Rat Hat. The reason being, "that apparantly the divers feel more secure when they can not see what is comming up behind them, then when they can", according to the safety manager [4].
In the Comex Hydra VI report [5] the divers summarize after 16 dives with the LBS equipment that, "the spherical shape of the window in the facial mask distorted the visual image and made it difficult for the divers to estimate distancies and dimensions". There is only one report dealing with theoretical analyses of possible distortion, "Design Demands on Diver Helmets"(in Norwegian) and our test method was designed on what seemed to be the most relevant tasks to perform [6]. Since visual survey of the surroundings depends on both field of vision and head movements, no perimetry tests were included.

EQUIPMENTS AND METHODS

Helmet
A spheric glass helmet was built for this purpose. It is well known that objects under water seems to be closer and larger than they are. This phenomenon depends on the difference in refractive index between the two media (water and air). When using a spherical glass a lens effect has to be added. It is to note that the spherical glass refractive index can be neglected, when calculating the lens effect, if the material is thin. The total effect can be described as follows:

Y' = L x ( R / (( n-1) x L + (n x R)) - ( t/L ) )

Where Y' is the apparent distance to an object.
L is the real distance to the same object.
n is the refractive index of the water ( normally 1.33 )
R is the radius of the spheres.
t is the thickness of the spherical glass.

Please note that the divers could easily move their head and the test thus only shows the differencies in performance caused by optical distortion, not a true difference in field of vision.

Fig. 1 Sideview of the test helmet and the SC 450 helmet

The sphere itself was made by 4 mm polycarbonate ( PC ). It had a radius of 200 mm and was connected to a collar of PVC. The total height of the helmet was 465 mm (fig 1). To fasten the helmet to the diver, a shoulder harness was used, and it was connected to the collar with three straps. The bouyancy of the helmet was compensated for by two leadpigs, one at the chest and one at the back. Their total weight in air was 35.8 kg.The helmet was supported by 80 l/min air through a hose, which was connected at the back of the collar, where the outflow took place. Inside the helmet the air was distributed by a T-shaped copperpipe, for demisting.
For reference a standard diving equipment was chosen, the AGA Divator mark II.

Test rig
The tests were designed to elucidate problems in distance and angular determination. In addition, a questionnaire on subjective opinions on the use of spherical and flat visor, was given to the test subjects.
The test rig consisted of a horizontal and a vertical wooden board fitted on a steel structure, placed at the depth of 5 meters in a well lit in door tank, measuring 6 by 4 meters. Upon the boards there were six teststations (fig 2). The teststations, numbered 1 to 6, can be described like this:

Teststation 1: A nine piece puzzle. The pieces had the shapes of different geometrical figures.Each piece will only fit in one place. The task was to to put them into their right positions and the time needed for this, was measured.

Teststation 2: A box of nuts and 21 bults. The task was to screw the nuts to a full thread. The time used was measured.
Teststation 3: Three holders for 590x105 mm paper strips, with 21 spots, numbered in such a way, that when hitting them in numerical order, the diver did this in a spiral. The task was to hit, with an awl, as many points as possible, with highest possible precision in a given time.

Fig 2. Test rig
Teststation 4: Resembles number 3, except that the holders form a V in the horizontal plane and there are 14 points designated A-N. They are placed so that the diver must change from left to right repeatedly, starting to the left near himself. Since the divers did not hit the same amount of spots, scooring in teststation 3 and 4 were made according to a weight function before statistical treatment [1].

Teststation 5: Seven sheets of paper, each having four almost identical figures. The task was to identify a correctly drawn geometrical figure on each sheet.

Teststation 6: A small weight, hanging just beneath the surface. The task was to direct and give orders for the lowering of the weight to a position straight over a small cup. The weight was hanging in an overhead crane and was never lowered. When the diver was satisfied with the position, the horizontal distance from the weight to the vertical axis through the cup was measured, in x and y co-ordinates, at the surface. While giving orders, the diver was standing on a fixed point on the bottom 3 m from the cup.

Divers
In the main test 37 different divers were used. Because of their various experience it was decided to divide the divers into four groups, two of experienced divers, "professionals"(>150 h) and two of less experienced divers, "amateurs"(<150 h). Two groups used the bubble helmet first, and two groups the full face mask, to eliminate possible effects of learning.
Each diver performed his two dives with a 25 minutes interval. This time was needed to change gear and to re-configurate the test equipment.

Statistical Methods
The results from teststation 1 and 2 were treated as pared data in a Student t-test. Since the results from teststation 3 and 4 were treated with a weight function the results were only used as a base for comparison [1]. The results from teststation 5 was treated with Wilcoxon Signed Rank Test, to test the hypothesis if there were any diffrencies between helmet and mask in identifying geometrical figures.

RESULTS AND DISCUSSION
Although there were great individual differences the mean values of the four groups in the Puzzle test (Test 1) and Nut test (Test 2) all indicated a faster accomplishment of the tests the second time it was performed irrespectively of the equipment. The differences in time between flat and sperical glass was in no group significant neither in the puzzle nor the nut test. Note in table 1, that the positive figures indicate that the second attempt was made faster than the first.

In the aiming test (teststation 3 and 4), track angles and scores showed that hits were placed toward the divers body when using the spheric glass while hits tended to occur away from the divers body when the flat glass was used (fig 3 ) .

The better results concerning the radius for the experienced divers when using the full face mask, can be explained by the fact that they are accustomed to optical distortion caused by flat front glass.

TABLE 1 Results from the puzzel test and the nut test, teststations 1 and 2
Order of visors Mean diff x ± s sec
Group of divers (n) 1st dive - 2nd dive Puzzle Nuts
"Professionals" 10 spherical-flat +2.8 ± 18.7 +7.9 ± 16.3
"Professionals" 11 flat-spherical +7.5 ± 15.9 +6.6 ± 9.3
"Amateurs" 10 spherical-flat +1.2 ± 32.9 +9.3 ± 11.6
"Amateurs" 6 flat-spherical +6.3 ± 10.9 +12.8 ± 20.4

A positive figure indicates that the second attempt was made faster than the first
When it came to the track angles, the results from diving with the spherical glass helmet, show the effects of its enlarging factor, and this was expected. On the other hand it was not expected that hits should tend to appear away from the body when the flat glass was used.However, no errors in this test was greater than 10 mm, with a fully streched arm, which is interpreted as good performance using both flat or spherical glass.

In the identification of geometrical patterns (teststation 5) no learning occurd since the divers were not given the right answers between the dives. The results show very small differencies between the equipments. Using the Wilcoxon Signed Rank Test, our hypothsis, that there is no difference between the flat and spherical glass, got a p value of 0.42, indicating no or a very small difference. The data are presented irrespectively of the order in which the visors were used.

TABLE 2 Results from the identifying test, teststation 5
Spheric glass helmet Full face mask

Answer alternative Correct Answer alternative Correct
A B C D Answer A B C D Answer Sheet
1 2 5 1 29 78% 3 4 0 30 81%

2 1 7 0 29 78% 0 4 2 31 84%

3 0 31 0 6 84% 0 28 0 9 76%

4 33 4 0 0 89% 37 0 0 0 100%

5 0 12 25 0 68% 1 13 23 0 62%

6 24 2 3 8 21% 12 1 3 21 57%

7 24 12 0 1 65% 20 16 0 1 54%
mean value 75.3% , s=±9.5% mean value 66.3%, s=±30.9

Correct answers are underlined
Spherical visor Flat visor
Fig 3. Results of the vertical part of the aiming test, teststation 3,
The test, in which guidence was given to an overhead crane operator, (teststation 6) gave an indication that the spherical glass helmet is superior to the full face mask, when it comes to tasks were wide field of vision is important. Unfortunately the enlarging factor makes it difficult to judge distances and the diver using spherical glass had almost twice the size in error which meant that they directed the weight to a point 0.4 m towards themselves, while divers with flat glass were 0.2 m closer to the reference point.

TABLE 3 Results of directing the weight, teststation 6
Mean value x ± s mm
Group of divers (n) Visors X Y
"Professionals" 21 spherical 58 ± 96 -441 ± 391
"Professionals" 21 flat 51 ± 60 -355 ± 252
"Amateurs" 16 spherical 26 ± 51 -321 ±289
"Amateurs" 16 flat 46 ± 87 -144 ± 336

Most divers placed the weight on their right hand side, this can be explained by the fact that the board and the cup were not placed in the middle of the tank, but slightly to the left hand side. The result of this could be, that if the diver uses the sides of the tank as reference and places the weight in the middle of the tank, it will end up somewhat to the right of the target.

CONCLUSIONS
No difference in performance was seen between flat and spherical visors. In general, distances are underestimated both in flat and spherical visors. However, it seemed as if professional divers performed less well when using spherical visor, probably due to their accustomness to flat visors under water. A magnifying factor was proved for the double curved glass, but on the other hand a fast adaption to the visor was observed. Only few negative comments were made on the spherical visor helmet, most of them related to the high boyancy of the prototype test helmet.
The results from the tests indicates that in most situations, there are no noteworthy differences between using a double curved glass and a flat one. One notable result was however found, namely that in exercises were a good survey is important, the double curved glass gave an advantage over an ordinary glass because of the better field of vision.
The radius is a critical factor and the distortion will increase if it is reduced. Further test will show if there is a smallest acceptable radius for diving helmets with double curved glass visors.

ACKNOWLEDGEMENT
The authers wich to express their gratitude to the divers of the Göteborg Fire Brigade, Naval District West and Chalmers Diving Club, without whom this project would not have been possible. We would further like to acknowledge the important assistance from the Department of Mathematics at Chalmers, in interpreting some of our results.

REFERENCES
1. Strannemalm, K. "Effects on Diver Performance when Diving with Double Curved Visor". Department of Naval Architecture and Ocean Engineering, Chalmers University of Technology, Report no X-90/5, 1990.

2. Hägglin, A., Bengtström, C. "SC 450 - A Total Concept Approach Airred at Divers Individual Equipment". UHMS Publication No 76 (UNDBR), 10/1/89, 1989.

3. Stuart, J. , Pirie, K. LBS System, Glenvarden Ltd, 74 Rosemont Place, Aberdeen AB2 4XS,Scotland, 1990.

4. Holland, D. Safety Manager, Oceaneering, Aberdeen, Telecom 1990.

5. Hydra 6, Comex, 1986.

6. Myrseth, E. et.al. "Design Demands on Diver Helmets (in Norwegian)". SINTEF STF23 F88001, 1988.


R t
Y' = L x ( (n-1) x L + (n x R) ) - L
Y' = Apparent Distance
L = Real Distance
n = Refractive Index of Water
R = Radius of Sphere
t = Thickness of the spherical glass


SC 450
Chalmers University of Technology