FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1919
1919 - 1078.PDF
been poured into the pails the pull on the stirrup was in agreement with the load-grading curve chosen. In order that the size of pail used would not be too large it was necessary to adopt three ratios of leverage for the levers. Beginning with the leading edge of the rib, the first six levers were made with a ratio such that the pull on the stirrup was three times the load in the pail. The next seven levers had a ratio of two to one, and the last seven, one to one. Operation of the Machine. The rib being assembled and secured to the dummy spars ready for test, 20 small grooved blocks were tacked along the lower edge as seats for the stirrups, to prevent crushing of the cap-strip and to simulate the sewing of the fabric to the rib, as done in the wing. The rib was placed on the platform of the machine, between the guides and the stirrups were attached to it. The stirrups were then attached to the levers, and the balance of the levers checked as a precaution. The four U-bolts were put on and connected to the steel cables leading to the floor, the lengths of the cables being adjusted so that all four would become taut at about the same time. The guides were fastened so that the rib would just slide easily between them. The purpose of these guides was to approxi mate the condition in the wing, where the ribs are cross-tied with tape and sewed through the wing every 2 or 3 ins. with a stout linen cord to prevent them from twisting out of the vertical. One round of cups or measures of sand of a fine, dry grade was placed in the loading pails to take up all slack in the wires, cables and stirrups. The total load on the rib at the end of one round was 41 lbs., distributed in this case according to the load-grading curve shown in Fig. 2. A steel scale was placed at the centre of the rib and a reading point selected and marked on the rib-web. A second round of cups was then poured into the pails, and a reading taken on the scale, for the purpose of comparing the relative rigidity of various'ribs. The loading was con tinued by rounds until failure occurred, the contour and the weak points of the rib being carefully observed meanwhile, together with the readings for rigidity. At stated points during the loading—as, for instance, at normal load, at two, four, six, or eight times normal load—photographs were taken of the rib. By comparing tracings, made from these photographs with the natural, free outline of the rib, an accurate measure of the rib contour under stated loads was obtained. When the rib failed, the point and probable cause of failure were noted. After failure, the amount of sand in each load pail was carefully determined, and the weight recorded with the pail number. From these weights, the total load on the rib was checked, and also the distribution of the load by means of a curve, plotted with the loading points as abscissce, and the pail weights as ordinates. This curve will have the same form as the load-grading curve, if the test has been properly carried out. The time required for test was three to five minutes, which was the disadvantage of using sand, instead of water. The water load can be applied in any time, down to about 15 seconds, by enlarging the orifices in the nozzles, or increasing the water-head. After test, the rib was removed from the machine, detached from the dummy beams, and weighed in fractions of an ounce. Progress in Design The method of experimental design is at best a " cut and dry " method. The first five ribs were designed by approxi mate mathematical methods. All the others were an evolu tion from them by making a study of the action of the rib during the test to destruction, and either making the weak section stronger, and the strong sections lighter, or by re designing the various details of the rib which failed, or showed extra strength, to secure the desired results. Ten types of ribs were experimented with, nine of which are shown in Fig. 4, the tenth being the well-known Handley- Page type of rib. Thirty-four ribs were made and tested before a satisfactory rib was produced. Since production has been in progress, a number of tests have been made which are not shown in the table. All the ribs tested were loaded according to the air loading shown in Fig. 2. This chart also shows an average typical loading, the result of a test, and gives a good idea of the small error in the loading by this method of testing. While this error may seem to be large at some stations along the chord of the rib, it is known by actual weight, and no interpolation is necessary to find the load at any station. The following table gives a complete analysis of this differ ence between the sand loading and the air load curve selected. The average difference over the stations is o. 78 lb. The aver age difference in per cent, is 2.7, and appears largely to be due to the loading at station 20. If we neglect station 20, the average per cent, of error is 1.8 per cent, instead of 2.7 per cent. Types of Ribs Tested The large table illustrates the various types of ribs designed - and tested, and gives a complete detail of each rib as to material used and the dimensions. r RIB TYPES /S|5^^S 1 - «<^EX 0=3= Fig. 4.—Various types of ribs tested. The reference letters are those used in the accompanying table Type " A " was designed along the standard type of rib generally used. It had a &-in. birch veneer web with the grain of the outer ply running horizontal. The cap-strips were -fa by £ in. selected spruce. Type " B " is the same as type " A," except that lein- forcements were put on the web to prevent buckhng. Ribs 2 to 8, inclusive, were of this same type and material, 7 and 8 being a little heavier than the others. These ribs also showed a strong tendency to bend in the horizontal portion of the rib at each portal near the vertical portion of the web. Type " C " is the same general design as type " B." The web in the centre, nose and tail sections being of *fr-ip.. thick spruce instead of ^-in. veneer. The " X " bracing was put on the web between each portal. In the tests this pre vented the shearing of the webs and leads to the next type. STAT/ONS I? 20 -Diagram of rib, showing stations and other items referred to in the accompanying table 1080
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
