COVER MATERIALS FOR TILE DRAINSMember C.S.A.E. an(j

F- R- Hore H. C. TiwariDepartment of Engineering Science, Ontario Agricultural College, Guelph, Ontario

performance of two types of glassfibre materials for use as drain filterswas evaluated by Nelson (1). Tileguard Type S-110 with random-reinforcing was found to be satisfactory,and a criterion was established bywhich it can be determined whethera drain in a given soil would be adequately protected by this filter material.

The purpose of this paper is toreport on a study of the performanceof Tileguard and Duramat, usedsingly and in combination with eachother, as cover materials for tiledrains.

MATERIALS AND PROCEDURE

Five different cover material treatments were studied in the laboratory. The five treatments studiedwere "blinding with topsoil" (check),"Tileguard above the drain", "Tileguard above and below the drain","Duramat above the drain", and"Tileguard above — Duramat belowthe drain". The experimental apparatus (Fig. 1) was designed to simu-

Silting of tile drains has been considered by many drainage engineersto be one ot the most serious problemsconnected with tile drainage of sandysoils. In Ontario there are many acresof poorly-drained sandy soils whichhave become highly productive inresponse to tile drainage. However,the life of the drainage system is usually short unless special protectivemeasures are taken to exclude theentrance of sand yet allow water movement into the drains. Common practices for this purpose have been towrap the upper half or two-thirdsof the tile joint with tar paper, orto blind the drain with straw, sawdust, shavings or other organic materials. Variable success has been experienced using these methods.

Recently two comparatively inexpensive glass fibre cover materials —Duramat (manufactured by GlobeGlass Saturaters Ltd., Petrolia, Ontario) and Tileguard (manufacturedby L. O. F. Glass Fibers Co., Toledo,Ohio, U.S.A.) have become availablecommercially. Tileguard is a glassfibre mat consisting of multiple layersof glass fibres in jack-straw arrangement held together with a phenol-formaldehyde binder. The mat israndom-reinforced by swirls of reinforcing yarn and is an average thickness of 0.020 inches. The glass is alime-borosilicate type. Duramat is abituminous-coated, parallel yarn reinforced glass fibre mat. The surfaceis dusted with inert material to prevent sticking in the roll and themat has limited porosity. The nominal thickness is 0.033 ± 0.003 inches.Sisson (3) reported that these materials have been recommended for useby several States in the United States.However, limited investigational workon their effectiveness has been performed. In 1959, Overholt (2) foundin a laboratory study that a glassfibre filter material was effective inreducing the rate of siltation and inincreasing the flow rate into a tiledrain. Glass fibre wrapped completely around the tile joint gave almostcomplete protection and was moreeffective than glass fibre wrappedover the top three-quarters of thetile joint. Sisson's (3) comparativestudy of seven cover materials showedthat a combination of glass fibreabove and plastic below was one ofthe best protective materials. The

1- Tile box 5. Constant-head tank2- Tile 6. Height adjustment screw3- Clamp 7. Control valve4. Manometer 8. Connection to Rainfall

Simulator

Figure I. General View of Apparatus Assembly

late field conditions around one tilejoint. Granby sandy loam subsoilwas packed uniformly into the bottomof a metal tile box 12 inches squareby 39 inches high at approximatelythe field bulk density. The subsoilwas scraped to the shape of a trenchershoe bottom. Two 4-inch diameterasbestos-cement tile were placed inthe simulated trench bottom so thatthe two outer ends protruded throughholes in each side of the tile box.The tile joint spacing inside the boxwas adjusted to i/8 inch and the tilewere clamped to guard against displacement during the backfill operation. The cover material treatment

was placed around the tile and therest of the box was filled with subsoil and compressed at a pressureof 1,000 lbs. A low rate of rainfallonto the soil was simulated andcontinued until ground-water flowconditions were established aroundthe drain. The rainfall was stoppedand ground-water flow into the drainwas maintained by means of a constant-head tank connected to inletsat the bottom of the box. With theassistance of manometers connectedto wells inside the box, water tableconditions were kept uniform for alltreatments by adjusting the height ofthe constant-head tank. This methodof applying water was chosen in orderto simulate a common field conditionwhere rainfall continues until a watertable in the soil is established. Thewater discharged and the soil movedinto the drain were measured by taking samples over successive 15-minuteperiods for six hours. By using thissampling technique, studies of variations in flow and soil movement during the test period could be determined. Mechanical analyses weremade on the soil moved into thedrain under each treatment. All treatments were randomized and replicatedfour times, and all results were analyzed statistically.

RESULTS AND DISCUSSION

Flotv of Water

The differences in accumulativeflow of water between treatments isshown in Figure 2. There was nosignificant difference between thetotal quantity of flow for the "Tileguard above and below" treatmentand the "Tileguard" above - "Duramat below" treatment. Differences inflow between all other treatmentswere significant. Examination of theflow curves showed that the dischargerate during the initial period wasconsiderably higher than the finalrate. This increased rate continuedfor approximately two hours afterwhich the discharge rate became almost constant. This indicates that thedrainage of gravitational water derived from the simulated rainfall persisted during this initial two-hourperiod and contributed water in excess of that being supplid by thesimulated ground-water flow system.The constant discharge rate towardsthe end of the test period was due

to the effect of ground-water flowconditions. A statistical analysis ofthe average discharge rates during thefinal two-hour period gave the sameresults as the analysis of the totalquantity of flow data.

Soil Movement

Figure 3 shows the quantity of soilmoved into the drain under eachtreatment. There was no significantdifference in the quantity of soil between "blinding with topsoil" and

§'*•••'• • itlilr •• ' f ,/y T,P:;'; \J'X-S • 1w',: ' y\ ' •%'•• •• f"• y jf1- :• • ....... ••• //T.. _. -"T~Bji:::: •;•::•..:. -. j?X •• H T|p:'— - - -yyA- : ^ -mil-; ;••:•.• •• /#• •••• y^ ^ V$i::: :li; :' /V- ••••••'••• y^ ••• ; y^-'jjj :•-!.:•; •• ^?* • .:•..• S ..• ' y'** -•pi|-; ; "4^ ''' •'- y'^- y^*' 'pii:; :; •': /^ S^yS". •• ! 'jjiii: -; | •:• : /J' ^"^^ , . 'pan : /p>7^ : ^•^•" 1MW^^-.r ' ": "^ J=s4-:"t 4 --••••••!•••:8 '̂;::&r ' ••' ' - .^ •—•""'"" 1 _i_ _|$P-7^U- y-"" •WL^ • -

K^-^^it-^^-^-------t-r^12 3 4 5 6

TIME - HOURS

Figure 2. Accumulative flow curves for differenttreatments.

.....__.._.

"223.3 - - — -

-

1>--

--

0^

^-

-

160

£

^^

11 — 113.9 —£ 1 - 2>|do £

-1^ ^

^... ^

Ss

^ - - 1P^ ^- 11# 16.9

# # 1 i %̂9.6

o 1—L %0 ^ s?1 6 §

- TREATMENTS -

Figure 3. Soil moved into the drain under differenttreatments.

—Key for Fig. 2 and 3—Tx—Blinding with Topsoil

T2—Tileguard aboveT3—Tileguard above and below

T4—Duramat aboveTr—Tileguard above, Duramat below

"Duramat above" treatments, nor between "Tileguard above" and "Tileguard above—Duramat below" treatments. Differences in soil movementbetween all other treatments weresignificant. Careful examination ofconditions around the tile made atthe end of each test revealed that therewas evidence of soil movementthrough the junction of the lower andupper layers of Tileguard under this

treatment. Soil movement into thedrain also occurred at the junctionof the lower and upper layers in the"Tileguard above—Duramat below"treatment. These observations indicate that wider strips of Tileguardshould be used on top. It was alsoobserved that under the "Tileguardabove—Duramat below" treatment thegroove in the simulated trench bottomwas completely damaged and loosesoil was present in the groove. Thisresulted in the slackening of the Duramat from the tile which, in the field,may lead to misalignment of tiles.Under the "blinding with topsoil"treatment, about four percent of thetotal soil moved into the drain entered through the crack space duringthe blinding process.

Under every treatment, all of thesoil movement occurred during thefirst hour of the test period when theflow of water into the drain was dueto the downward movement of gravitational water from the soil above the

drain plus equilibrium ground-waterflow conditions. This observation em

phasizes that in the field soil movement will occur during and shortlyafter a rainfall rather than solelyunder equilibrium ground-water flowconditions.

Significant changes in the composition of the original soil and the soilmoved into the drain occurred only

_I00

| 90

1 1 1 ! 1

1 ^ iORIGINAL SO

TREATMENTL ff

£ TO

// (/

8! 50

130

i /

/

/

| />-"'

"' 1S^""'" 1 :

Figure 4. Composition of the Original Soil and theSoil Moved Into the Drain (Blinding with TopsoilTreatment).

11 1 1 i! ! : 1 t^T"

! ! ! • • iT 1i j//

; 1

77I'/

/1 ! ! : i i ; !

! i ! i i 1 \y\ 1 iI : i ,--'f

r i i ,>-'! i/ !M^^n^l 1 i

500 1000 2000

Figure 5. Composition of the Original Soil and theSoil Moved Into the Drain (Tileguard Above andBelow Treatment).

under "blinding with topsoil" and"Tileguard above and below" treatments. The composition of each soilis shown graphically in Figures 4 and5. There was a significant increase of

29 percent in the sand fraction anda decrease of 40.2 percent in the siltand clay fractions under the "blinding with topsoil" treatment. It is apparent that some subsoil as well astopsoil was carried into th drain. Inthe "Tileguard above and below"treatment there was a significant increase of 185-1 percent in the silt andclay content over the original soilcomposition. The increase in fine particles is explained by the fact that thecrack was covered almost completelyby an effective filter material. Thecoarser particles that did enter thedrain probably passed through thejunction of the lower and upper layers of Tileguard.

SUMMARY

Five cover material treatments werecompared to determine their effecton the movement of soil and waterthrough the tile joint into the drain.The experiment was designed to simulate field conditions around one tilejoint as closely as possible.

Based on their performance, thetreatments can be rated in the following order:

Tileguard above and below

Tileguard above—Duramat below1.

2.

3.

4.

Tileguard above

Duramat above

5. Blinding with topsoil.

It was observed that soil movement

into the drain occurred due to gravitational water flowing from the backfill material above the drain and the

ground-water flow conditions hadcomparatively little effect on soilmovement.

The results of this study suggestthat further research should be con

ducted on treatments using widerstrips of cover material above thedrain. They also indicate that theeffect of more cycles of alternate rainfall and ground-water flow on soiland water movement should be

studied.

REFERENCES

1. Nelson, R. W., Fiberglass as aFilter for Closed Tile Drains.

Agric. Eng. Vol. 41: 690-693, 700,1960.

2. Overholt, Virgil. Fiber Glass Filters for Tile Drains. Agric. Eng.Vol. 40: 604-607, 1959.

3. Sisson, D. R., A Comparison ofSome Filter Materials for TileDrains. Unpublished M.S. Thesis,Univ. of Illinois, Urbana, Illinois,1960.