Professional Documents
Culture Documents
INTRODUCTION
The methods of Burmister and Southwell were used in an earlier Paper (Fox 1948)r to calculate stresses, in a two-layer system, arising from uniform loading applied over a circular area at the top surface. The relaxation method was used, in a few specific cases, to give an indication of the spread of stress while Burmister s method provided accurate values, for a much wider range of parameters, of vertical and horizontal stress at any point on the vertical axis. In the two-layer system (Fig. l), two Fig. 1 parameters are involved. These are the ratios : ai = a/h, and k = 3
1
Layer I
11111111 lJlillll
-2. . c
E2
where a denotes the radius of the loaded area, h the thickness of the top layer, and E,, E, are respectively the moduli of elasticity of the upper and lower layers. Poisson s ratio was taken to be 0.50 in both layers. Two types of condition at the interface, given by (a) no slipping, and (b) no friction, were considered ; the true state
1 Fox, L., 1948. No. 9. Computation
El
T
Layer 2
I
293
EZ
D.S.I.R.,
Pajw
294
W.
E.
A.
ACUM
AND
L.
FOX
of affairs probably lying nearer to the former than to the latter. were performed :L (1) Burmister s method (a) No slipping..
2 and Gs (radial stress in lower layer) at interface on vertical axis, for al = a, 1, 2, 4 and k = 1, 2, 5, 10, 32.3, 100, 500, 10,000.
E and .Z-
5s at points on vertical axis in lower layer, for aI = ), 1, 2 ; and K = 1, 10, 100, 1,000. .Z and .G - Gs at points on vertical axis in lower layer, for al = 4, 1, 2 ; and k = 1, 10, 100, 1,000.
(b) No friction.
(2) Relaxation method (a) No slipping. Full stress picture, for al = 1 ; and k = 1, 2, 10, 100. This article deals with the extension of the analysis to the three-layer case.
EXTENSION TO THREE-LAYER CASE
The extension of this work to the three-layer system, promised in the earlier Paper, has now been completed, and the relevant formulae and results follow. Unfortunately the application of relaxation methods to this case was found to be impractical, the amount of work involved to secure reasonable accuracy being prohibitive. Results are therefore available only for stresses on the vertical axis, but it is considered that the two-layer relaxation results give a satisfactory estimate of the off-axial spread of stress. The extension of Burmister s method is practicable, though the number of parameters has increased to four, and the expressions for the stresses have become correspondingly more cumbersome. The chosen parameters (Fig. 2) are given by aI = a/hz,H = hl/hz, k, = El/E,, and k, = E,/Es,
Fig. 2
111111! -2aT
llllll
I I
and the following stresses have been computed on the assumption of perfect friction at each interface, and with Poisson s ratio equal to 0.50 throughout. (1) At first interface
The chosen values of the parameters are as follows :ar H k, k, = = = = 1, 2, 5, 1, 4. 1, 4, &. 10,20,50, 100, 500. 5, 10, 50, 100.
STRESSES
IN
A THREE-LAYER
SYSTEM
295
Some alleviation of the complicated nature of the stress formulae, given below, arises from the fact that the parameter al enters somewhat trivially into the formulae. (1) First interface
2 = (1 2za= -(l
fw - K)
FI Odx N) o kmx)D(x)
=LJ&rx)G odr s 0 D(x)
OD o%fd GA4 UX-dX 1 )D&)
s
4)
2's=
t1 -K)(l
The symbols
and functions
in these expressions
k, - 1
K=kl+l
fi(x) =
N=k2+1P&z--~~~ 1 N2e-13JJ+4jz 1
(1 + Rx)
{ -
e-J-rz-- KNse-(4+mZ
1 _
(1 - Hx)
~$l(x)e-(3H+2kC
Nfi2(x)e-UZ+2k
a(x) =
i--&(1-
Hx)(,,3,
- N2e-Qa+4)2}-
A--(1 + Hx){e-Hz
Ke-3EZ
_ jjle-C3lZ+4)Z]
K(2 - 5Hx)e-aHZ -
W%2)e-~+3)z x _ Hx)e-CJH+B)z
+ 2x2)e-ss + KWee-4z
=(I + m2,2),y2Zh
+ K2e-uz
+ 2KN(l.+
_K~N$,(X)~-I~I?+~)Z
W2x2)e-f2R+4k%
W.
E.
A.
ACUM
AND
L.
FOX
+2x-H%
+2%* -2HXe-4Hz3)+(1-2z-H~x+2S+W~~)
pl( - x) = K(4 + 2Hx + 8x8 + 4Hx3) - (2 - 4x - 5Hx - 2~2 - 2HXn) - Ks(2 + 4x - 5Hx + 109 + 2Hx= - 2OHxS). K(2 + 4x + Hz - 2x2 - lOHn% + 4Hx8) - (2 - 4x - 5Hn - 2x0 . + 4HW + 16HW. + @HS + 4HW. 1OHS) - (2 - x - 5Hx + 2Hx*) (2 - x - Hz) (2 - 5x - 5Hx)
= $e~z)=~)
PRESENTATION
OF RESULTS
In the two-layer case results were given in the form of tables, graphs, and contours. When four parameters are present, however, it is very difficult to cover the whole range adequately and to present the results in a suitable manner. The results are therefore given in tabular form only, and it is left to the user to plot them, if desired, in any convenient way. The wide range both of arguments and results suggests that plots on logarithmic scales will be the best. The vertical stress Z is always of one sign, and can be plotted in this way without difficulty. The horizontal stress F7 is of variable sign, but the more important quantity ,rz - $7 is of one sign throughout, and it is therefore more convenient to plot ,Z - f? rather than G?. In Tables 1 to 8 (Appendix), stresses are given as percentages of the applied load. All the signs are negative, and the six values given in any cell of the tables are arranged according to the following scheme :First interface Second interface
The vertical and horizontal stresses were computed independently, and an almost perfect check, apart from a constant &merical factor, is given by the relations
az,+?I =R1(21 fi2) at
first interface,
and Special cases o&r the two-layer system. in the previous work.
(1) Stress req&ed
Es - Es = ka(Zs - F?s) at second interface. when Kr or kg becomes unity, and the problem is reduced to that of Some results can then be obtained by interpolation or extrapolation There are four special cases.
atfirst interfuce, k, = 1.
This requires a knowledge of the stress at points in the upper layer of a two-layer system, for which no computations have been performed.
STRESSES
IN A THREE-LAYER
SYSTEM
297
atJirst interface, kg =
1.
This corresponds to the det ermination of stresses at the interface of a two-layer system where, with an obviousnotation, the parameters are given by /a,@ = @)/II, (3) Stress re+ired at second in&face, kl = 1. This corresponds to the determination of stresses at the interface of a two-layer system, with parameters w = & s. /(l + H), and k ur n = @ (4) Stress required at second iderface, k, = 1. This corresponds to the determination of stresses at a point in the lower layer of a twolayer system. The parameters are al(s = al s/H, and k s = kl@, and the depth below the interface is k s/H. No values are given in the Tables, since it is generally easier to obtain them, if required, by extrapolation of the log-log plots of the three-layer results. Enough values were obtained from the two-layer work, however, to prove that there was no signXcant error in the complicated formulae and computation. and k s -= kl a.
ACKNOWLEDGEMENTS This article has been written and the results included therein have been obtained at the Mathematics Division, National Physical Laboratory, as a part of the programme of the Road Research Board ; it is published by permission of the Director of Road Research.
$1
9.51
.I
3.64
10
6.43
0.552
298
W.
E.
A.
ACUM Table
AND 2
L.
FOX
H = l,a,
= 1
k, 4
5 30.30 162.00 32.30 27.90 173.00 34.70 5 9.22 37+3O 7.56 6.11 47.10 4.70 20.80 229.00 22.90 18:70 250.00 25.00 10 7.14 28.10 5.63 4.73 35.00 3.50 13.50 300.00 15.00 11.80 330.00 16.50 399.00 20 5.33 19.70 3.96 3.53 24.40 2.44 7.26 390.00 7.84 6.13 431.00 8.66 50 3.44 11.40 2.29 2.26 13.90 1.39 4.43 457.00 4.57
L
100 2.36 7.10 1.42 1.38 594.00 1.19 1.04 645.00 1.29
10
:t:::
100 --
201.00
68.20
298.00
50.90
It
1.55 8.55 0.86 0.572 11.80 0.236
4 k,
5 66.40 215.00 43.00 63.40 223.00 44.60 59.80 239.00 47.80 59.10 244.00 48.80
20
15.60 62.00 12.40 10.40 78.00 7.81 39.90 561.00 28.10 37.00 615.00 30.70 33.50 730.00 36.50 32.80 776.00 38.70 12.90 49.80 10.00 8.59 62.80 6.28 3.14 93.30 1 s7 24.70 868.00 17.40 22.20 963.00 19.30
50 9.42 34.00 6.81 6.29 42.90 4.29 2.34 63.90 1.28 1.51 16.20 1116.00 11.20 14.20 1244.00 12.40 11.60 1541.00 15.40
100 7.06 23.80 4.77 4.72 29.80 2.98 17.70 44.20 0.88
500 5.48 1687.00 3.37 4.43 1866.00 3.73 3.13 8.60 1.72 2.08 10.50 1.05
10
50
100
1 I
3.74 115.00 2.30 19.10 1 180.00 23.60 18.50
1
!
/
,
Table
H =
a, a,
--I
kl
I
5
&I
I I
I
1
10
I614.00
90.70
1273.00
75.20
1966.00
61.20
3991.00
3il.i
657.00 j:::: 90.20 70.80 14.20 6.52 197.00 3.94 85.50 30.20 5.91 3.69 76.40 725.00 36.40 75.80 745.00 37M
/ :::::
115.00
1 :::::/
1404.00
95.60
::::
/ :::::I
2195.00
77.90
S::: 1 l::::i
4500.00
3810;)
;.::
I 303.00 I 185.00
STRESSES
IN
A THREE-LAYER Table 5
SYSTEM
299
H = 2, a, = ;
k, k,
5 2.85 21.50 4.30 2.51 23.10 4.62 2.12 27.10 5.42
5 0.950 4.09 0.817 0.621 4.96 0.495 0.220 6.80 0.136 1.76 27.20 2.72 1.52 29.40 2.94 1.23 34.60 3.47
10 0.670 2.70 0.540 0.435 3.24 0.323 0.155 4.40 0.088 1.06 32.40 1,62 0.884 35.00 1.75 0.682 41.20 2.06 0.644 44.10 2.21
20
50 0.268 0.853 0.170 0.307 42.10 0.421 l0.241 45.20 0.454 0.064 1.32 0.026 0.165 53.70 0.537
500
l-
10
50
Table Ii k, k, 12.00 66.10 13.20 11.30 69.10 13.80 10.50 76.10 15.20 IO.30 79.00 15.80 5 2.55 10.90 2.18 1.66 13.30 1.33 0.586 18.70 0.374 0.370 21.10 0.210 7.72 87.10 8.69 7.13 92.30 9.22 6.46 104.00 10.40 6.32 109.00 10.90 10 1.93 7.89 1.58 1 a26 9.68 0.969 0.453 13.70 0.275 0.287 15.40 o-154 4.74 107.00 5.36 4.29 115.00 5.74 3,74 132.00 6.61 3.64 140.00 6.99 20 1.42 5.43 1.09 0.929 6.64 0.664 0.336 9.39 0.189 0.215 IO.60 0.106 Table =
1, a, = &
50 2.39 132.00 2.64 2.08 142.00 2.84 1.71 165.00 3.30 1.64 175.00 3.50 7 0.899 3.07 0.616 0.586 3.71 0.371 0.213 5.14 0.103 0.137 5.75 0.057 1.40 149.00 1.49 1.18 160.00 1 a60 0.927 186.00 1.86
100 0.610 1.89 0.376 0.400 2.26 0.227 0.145 3.08 0.062
500 0.404 184.00 0.368 0.318 197.00 0.394 0.235 os550 0~111 0.152 0.637 0.064
10
50
100
4 kz
9.41 363.00 7.26 8.71 388.00 7.76 34.20 151.00 30.10 33.90 152.00 30.30 I,20 39.60 0.794 0.748 44.10 0.44 1 7.88 443.00 8.87 7.72 466.00 9.32
50
2.58 9.69 1.94 1.69 12.00 t 1.20 0.612 17.40 0.348 5.76 433.00 4.32 5.20 465.00 4.65 4.51 54: 40: .
100
1.89 6.63 1.32 1.25 8.18 0.818 0.457 11.80 0.237
50
100