METHOD OF TESTING DRAIN TILE.

BY

ARTHUR

N.

TALBOTt AND

DuFF

A. ABRAMs.t

Recent developments in the manufacture of farm drain tile

have emphasized the importance of having a simple standard
portable testing machine which may be used for making tests
of drain tile in the field or at the plant. The increasing use of
tile of large size in farm drainage districts is well known. The
competition between clay tile and the concrete tile has brought up
new questions. What strength shall be required for tile of a
given size in order that they may be considered to be commercially
first-class tile? In the case of concrete tile, what thickness,
richness of mixture, method of curing and age at laying are necessary to fill the requirements for a first-class article?
A number of elements enter into the choice of a suitable
niethod of determining the physical properties of drain tile:
(1) A definite and important quality should be determined
by the test.
(2) The test should be simple, easily and quickly made and
should not require the services of an expert laboratory man.
(3) The test should be of such a character as not to give
unduly diverse results for test pieces of the same grade.
(4) The machine to be used should be simple and inexpensive,
readily adjusted to different sizes of test pieces and easily transported from point to point and made ready for use.
It is believed that the machine described in this article
satisfactorily fulfills the requirements for making field tests of
drain tile. This machine was designed by D. A. Abrams for use
in the Laboratory of Applied Mechanics of the University of
Illinois.
The machine consists essentially of a simple frame-work and
a lever for applying the load by means of dead weight. The load
* A part of the material in this paper was presented at the annual meeting of the Interstate Cement Tile Manufacturers' Association at Chicago, February 22, 1911.
t Professor of Municipal and Sanitary Engineering, in charge of Theoretical and Applied
Mechanics, University of Illinois.
~Associate, Engineering Experiment Station, University of Illinois.
(713)

714

TALBOT AND ABRAMS ON TESTING DRAIN TILE.

applied through the loading lever may be blocks of iron, stone,

sand or other suitable material. After the test the dead load is
weighed. To obtain the load on the tile this weight is multiplied
by ten and a constant quantity due to the weight of the loading
lever (about 100 lb. in this particular machine) is added.
The machine with a 30-in. clay tile in place ready for loading
is shown in Fig. 1. Fig. 2 gives the principal dimensions of the
different parts. The machine measures 30 in. between uprights
and will take tile up to 42 in. inside diameter. The main members

FIG. 1.-T!LE UNDlm TEST.

are of timber; metal plates and other shapes are used at points
of concentrated load and for connections,
Metal knife edges are provided for the bearing of the loading
lever on the top loading block and for taking the upward thrust
against the top cross block. The knife-edge bearings on the block
over the test tile are 5 in. center to center and a single knife edge
takes the end thrust. This gives considerable freedom to the
top loading block and allows the load to be fairly central, although the top and bottom elements may not be parallel,

TALBOT AND ABRAMS ON TESTING DRAIN TILE.

715

The bottom loading block is provided with two small halfrounds of hard wood placed about 2 in. apart, which allow the
tile to seat itself in place. The load is applied at the top along
a single element. Cushions consisting of short lengths of flattened
rubber-lined fire hose serve to distribute the load along the length
of the tile and prevent any local concentration of the load due to
irregularities in the top or bottom surfaces.
The top cross block can be placed and held in any position
along the uprights to accommodate the machine to any diameter
of tile up to about 42 in. By this means the machine is adjustable

:.os~o"

~-r-Jt~A

l:

li !
i..iU/

Lll
F"ron+

5ide
I fl.

FIG. 2.-DETAILS OF TESTING APPARATUS.

to the greatest variation in the size of the test tile and will apply
the load to any sizes under uniform conditions.
In order to check the dimensions of the loading lever, it was
calibrated by setting a pair of platform scales in the machine and
loading up to about 500 lb. on the machine. It was then placed
in a 10,000 lb. testing machine and loaded up to 4,000 lb. The
greatest error observed for this range of load was less than 1 per
cent.
This machine weighs 225 lb. It should not cost more than
$15 to $18 in a shop equipped for wood and metal working.
Up to the present date abo11t op.e huud:red tests hfl,ve been

716

TALBOT AND ABRAMS ON TESTING DRAIN TILE.

made on this machine on concrete and clay tile in sizes 12 to 36 in.

inside diameter. The breaking loads varied from 1,400 to 5,000
lb. per tile.
An examination of this testing machine wi.ll show that it is
simple in operation and that it is easily adjustable for different
sizes. Tile which are out-of-round in different ways at the two
TABLE I.-SUMMARY OF TESTS OF CoNCRETE DHAIN TrLic.
~--~-

Ref.
No.
----

Average
Internal
Diameter,
in.
-~-

Average
Modulus
Thickness, Length, Weight, Age at Maximum
of
Top and
Test,
Load,
Rupture,
in.
lb.
Bot,tom,
days.
lb. per
lb.
m.
sq. in.
--- --- - ----

Remarks.

---

---~

36W1
36W2
36W3

36.7
3tl.O
36.0

2.95
2.98
3.02

23.8
24.0
23.9

tl90
700
722

177
177
177

24W1
24W2
24W3

23.9
24.1
24.0

2.07
2.00
1.97

24.0
24.0
24.0

313
312
316

181
181

20W1
20W2
20W3

20.3
20.3
20.3

1. 74
l.H3
1. 70

23.9
23.9
23.9

205
209
209

163
159
159

18W1
18W2
18W3

18.3
18.3
18.3

1.05
1.05
1.05

23.8
23.8
23.8

188
186
184

13T>

12W1
12W2

11.8
11.8

.95
.96

12.2
12.2

30
31

14fi
146

13:i

135

4,310
4,680
4,760

770
819
812

Average

800

2,860
1,900
2,660

690
492
708

Average

615

1,770
1,640
1,670

515
539
516

Average

527

3,000
2,920
3,170

882
8.19
892

Average

878

850
820
Average

See note below

939
887
913

I
-----~--

The 36-in. tile were reinforced with 2U-in. square twisted bars, placed at middle of thickness of tile. Bars were welded into circular rings. The modulus of rupture for these tests was
computed in the san1e manner as for the other tile, disregarding the reinforcement.
Concrete consisted of 1 part Portland Cement to 3~ parts washed sand and gravel.
Concrete machine 1nixed.
The 12-in. tile were mcwhine rnude and were placed in steam chamber for 12 hours. The
other tile were stored in the open air.

ends will be easily taken by the machine and there is little chance
for an unfair distribution of the load. The strip of hose gives
some cushioning effect and the load is practically distributed over
the whole length in all cases. The method of loading along a
line at the top and bottom was selected because of its simplicity.
The arrangement of the machine allows a tile to be rolled into

TALBOT AND ABRAMS ON TESTING DRAIN TILE.

717

place and to be easily made ready for test. It is believed that

the results obtained by different operators will agree quite closely.
If desired, the modulus of rupture of the material may be
determined from the bending moment developed and the dimensions of the pipe. For general purposes it will be preferable to
report the load per foot of length of pipe for a given size. Possibly
for some purposes it may be interesting to divide this load by the
TABLE II.-SuMMAHY oF TEsTH oF CoNcnETJC DRAIN TILE.

Ref.

No.

Average
IutPrnal
Diameter,
in.

;':yerage
1\Iodulus
1hJCkucss, Lt>ngth, W 1igltt., ~gf' at ~Iuximtllll
of
Top aud
in.
lb
fest,
Loud,
Rupture,
Bo~tom,

1n.

days.

lb.

lb. per
sq. in.

l,9(j5
1,885
2,495

314
304
451

36-1
36-2
36-3

35.9
36.0
36.0

3.10
3.10
2.93

24.3
24.2
24.0

706
721
700

101
72
105

27-1
27-2
27-3

26.7
26.6
26.6

2.81
2.90
2.84

24.3
24.6
24.5

454
460
475

65
84
60

22-1
22-2
22-3

22.0
22.0
22.0

2.24
2.23
2.27

24.4
24.1
24.2

321
335
315

51
68

12-1
. 12-2
12-3

12.0
12.0
12.0

1.05
1.05
1.05

12.3
12.2
12.4

39.6
39.3
39.3

Average

356

2,135
2,945
2,305

311
411
327

Average

350

2,fl(j5
2,885
1,890

500
557
350

Average

1,405
1,295
1,345
Average

Remarks.

Wet concrde

Wet Concrete

469
1,290
1,200
1,225

Machine made
do
do

1,238
. - - -.. - - - - -

The 36, 27, and 22-in. tile were made at a field plant about two miles west of Champaign,
Illinois. Metal forms were used. The forms were removed immediately upon the completion
of the tamping.
The 12-in. tile we1e machine made, having been shipped from the plant at Edinburgh, III.
These tile had 6 circumferential corrugations about 0.10 in. high and t in. wide at base.
Details regarding materials used and storage conditions not furnished.

diameter of the pipe in inches and thus compare the results per
inch of diameter for a pipe one foot long.
It has seemed the simplest way to fix at a definite distance
apart the two strips on which the tile rests. An analysis of rings
shows that when the bearings on these strips are 2 in. apart, the
formula for the bending moment will be but 2! per cent different
from that for a single support for tile 6 in. in diameter and l of 1
per cent. for a tile 12 in. in diameter, while for larger sizes this

718

TALl30T AND ABRAMS ON TESTING DRAIN" TILlll.

variation will be much less. Under the conditions of such tests

it would seem better to fix the distance for these strips and use a
common expression for the formula for the bending moment for
all sizes of tile to be tested. It would seem that 0.16 Qd is a satisfactory expression for the bending moment where Q is the concentrated load applied at the crown and d is the mean diameter
of the tile. For the modulus of rupture f of the material the
formula would be
Qd

/=0.96 lt

where l is the length and t the thickness of the tile along the top.
This method of testing was selected in preference to a method
involving the bedding of the tile in sand or other material, because
of the difficulty in embedding large tile in sand in such a way as
to obtain a fair distribution of pressure and in securing the same
distribution of pressure in different tests and because the method
of concentrated loads will give a more definite index of the strength
of the material.
In tests of materials it is not essential that the material shall
be subjected to the same action in the process of testing that it
will receive in service. The cold bend test of steel is one of the
most useful and instructive of tests, but it differs radically from
any condition of service in which the steel will be placed. The
value of a test will depend upon the properties determined. In
testing drain tile the method of applying concentrated loads
has many advantages over that of applying distributed loads.
Whatever the method of testing used, it will be necessary finally
to determine the relation between strength of the test piece and
the strength which is needed in the structure. In the case of tile
to be used in a ditch of a given depth and a given soil the necessary
test strength will have to be determined. Since the tests will have
to be translated into the working conditions, it would seem
unnecessary to attempt to make the conditions like the conditions
in the ditch. It is of much more importance that the tests should
be simple, direct and fairly uniform under varying conditions of
tile and with different machines and different operators. Experience with this machine leads to the conclusion that it would make
a satisfactory means of determining the quality of drain tile.

TALBOT AND ABRAMS ON TESTING DRAIN TILE.

719

The three series of tests of drain tile given in Tables I, II

and III have been selected as representative of the results obtained
with the machine described above and were made at the Laboratory of Applied Mechanics, University of Illinois.
TAHLE 111.-SUMMAHY OF TESTS OF CLAY DHAIN TILE.

~':'erage
1

Ref.
No.

Average
I Internal hwkness Length,
Diameter, fop and
in.
in.
Bo~tom,
1

-30-1
30-2
30-3

-~~

30.5
:l1J.2

2.30
2.35
2.35

27-1
27-2
27-3

28.0
2!LO
27.6

24-1
24-2
24-3

,
:

.
11\!adulus
Weight, Maxmnnn.
of
lb.
Load,
Rupture,
lb.
lb. J!er

~1----:;;24.0
24.0

416
406

1.90
1.97
1.93

25.5
24.5

:l08

23.5

313
314

24.6
24.6
24.5

1.78
1.71
1. 72

25.5
25.4
25.4

238
241
242

18--1
18-2
18-3

18.9
19.0
19.0

1.30
1.32
1.30

25.8
25.2
25.5

145
144
143

12-1
12-2
12-3

12.7
12.4
12.6

1.10
1.07
1.12

25.8
25.2
25.5

79.7
80.0
79.3

Renmrks.

~m.

4,850
3.750
2,830

Average

-;,;~~--~
905
660

\188
1,066
1,180

Average

1,045

2,360
2,.520
3,030

737
818
1,or,5

2,040
2,080
2,240
Average
1,835
2,985
1,745

Hard burned
Black core
Black core

915

3,180
3,640
3,840

IAverage

870
917
915
1,025
952
760
1.318
713

~ _ _ _ __

- - - - ' - - - - - - - - - ' - - - - ' - - - - - Averag: _ _