NOTES ON RIVETS AND SPACING OF RIVETS FOR OIL-TIGHT WORK. 81 
due evidently to the seam on the inside being closed by the greater deflection. All the 
leakage was through the butt caulking in the seam of the 12-pound plating. A few rivets 
showed signs of leaking, but leakage was too small to measure. Electric welding and 
other connections were perfectly tight. Twelve-pound side of the tank deflected about 
1 inch, the permanent set being 14 inch. 
““65-FOOT HEAD OIL TEST IN SECTION (B). 
“This section of the tank was tested and pressure gradually applied similar to section 
(A) with very poor results. The butt caulking in the single seam of the 12-pound plating 
leaked continually, andit was necessary to refill the stand pipe five times during the test. 
The riveting and other connections on the section of the tank showed up very good with 
the exception of a few rivets which showed some signs of leaking, too small, however, to 
measure. The deflection and permanent set were about the same as deflection in section 
(A). In view of the fact that the seam in the 12-pound plating leaked so badly, it was 
not considered advisable to subject this section to the gasoline test. 
‘“45-FOOT HEAD GASOLINE TEST ON SECTION (a). 
‘“The tank was removed from the oil house to a vacant space where danger of fire 
would be minimized. The tank was filled with gasoline until the stand pipe overflowed. 
The pressure was gradually applied until the gauge registered 20 pounds. Seam in 12- 
pound plating leaked as in other tests, beginning to leak shortly after the tank was filled. 
Efforts were made to get an accurate account of the leakage of gasoline, but this was 
impossible due to the gasoline evaporating. The amount of leakage caught during the 
twenty-four-hour test under the 20-pound pressure was g pints. All the leakage was 
through the above-mentioned seam with the exception of a slight leak around the toe of 
the bounding bar where same was riveted to the 15-pound plating. 
““65-FOOT HEAD GASOLINE TEST ON SECTION (B). 
“After completion of the twenty-four-hour test under 20-pound pressure the pres- 
sure was gradually raised to 30 pounds. The leakage was slightly greater than under 
2o-pound pressure but in the same place. Total leakage during test under 30-pound 
pressure amounted to 334 gallons. There were a few rivets that showed slight signs of 
leakage; leakage, however, was too small to measure. 
“This tank as a whole showed up well under the various tests, with the exception of 
seam in the 12-pound plating. To overcome this defect it was decided to renew the 
12-pound plate, using a lap instead of a single strapped seam, rivet-spacing the same. 
The tank was again tested with water, air, oil and gasoline, using 30-pound pressure as 
the maximum. Both compartments of the tank were perfectly tight with water, air 
and oil. A slight leak developed between the bounding bar and the 15-pound plating in 
section (A), while it was under a 30-pound gasoline test; that is, after the 30-pound pres- 
sure had been on for about an hour, pressure then dropping 2 pounds in six hours, leakage 
¥% gallon. 
““Gasoline test pressure of 30 pounds was applied to section (B) similar to section 
(A) and allowed to remain on for six hours. A bad leak developed about three hours 
after the 30-pound pressure was on between the inner seam strap and the 15-pound plat- 
ing, the leakage entering section (A) about two feet from the division bulkhead. Both 
corners of the bulkhead bounding bar, where same are riveted to 15-pound plating, 
