The edge crush test

William 0.Kroeschell

The choice of ECT as the measure of strength in the new AlternateJi.eight reguhtionsprovides a
naatev-ial standdrd much more closely rehted than burst to containerp4omnce.

The edge crush test (ECT) or short column test for corru- most always occurs. At the time of McKee's work, prepara-
gated board is in the limelight today for two reasons. First, tion of the necked-down shape was tedious. McKee had noted
this test has been adopted by the railroad and trucking classi- that reinforcement of the loaded edges of a rectangular speci-
fication committees as an alternate to bursting strength in men was also effective in eliminating premature failure at the
determining the strength of corrugated board used in ship- loaded edge. It remained for Koning, also at the Forest Prod-
ping containers-reason enough for a better understanding ucts Laboratory, to q u a n t a the relationship and conclude
of the test. The second reason is that ECT is directly related that edge reinforcement of the rectangular specimen was
to box compression strength and, in today's shipping environ- preferred (5).Recently, however, Koning has devised a new
ment, box compression is the most important factor affecting means of cutting the necked-down specimen and he now
box performance. As we will discuss,we cannot ignore the fact prefers this technique (6).
that ECT is not the only factor affecting this performance. If
we focus entirely on ECT we risk box failures. This could Test preparation
result in withdrawal of the permission to use the advantages
of the Alternate regulations. There is further interest in edge Many researchers have identified the fact that the method of
crush because of the discussions in the international commu- cutting the test specimen is critical, although with edge rein-
nity toward finding an international standard test method. forcement this may not be as critical as was reported by the
many who worked with unreinforced pieces. In a more recent
Background study, Schrampfer and Whitsitt (7) in testing unreinforced
specimens, also at the IPC, found that specimens cut with the
The edge crush test that is to be used in determining compli- circular saw with the sharp hollow-or taper-ground blade
ance with the new Alternate Rule 41/Item 222 is described by prescribed in T 811 gave the highest test results. However,
TAPPI Test Method T 811 (1). Briefly, this method pre- the use of a double-bladed motorized cutter with sharp single-
scribes a test specimen 2 in. wide with varying height, de- bevel blades yielded results so slightly lower that, with edge
pending upon the flute construction. The loaded edges of the reinforcement, the differences may be insignificant. They
specimen are reinforced by impregnating with paraffin wax. have stated that this alternative procedure may be an accept-
The test is performed in a compression machine. able substitute for the saw specified in T 811.The single-bevel
blade operated with the bevel facing away from the test
Edge reinforcement specimen appears to be critical in this alternative. It is also
important that the test specimen be cut with the loading
The development of the method and the reasoning behind the edges exactly perpendicular to the flutes (8).
various aspects of the technique have been described by
Kroeschell(2). The first researcher speaking for the impor- Testing machine
tance of the test was Kellicutt of the U.S. Forest Products
Laboratory in Madison, WI (3).The basis for the T 811 test The test originally proposed to the carriers for incorporation
was the work of McKee and his associates at the Institute of in the Alternate Rule 41/Item 222 was TAPPI T 823, which
Paper Chemistry (IPC)", in Appleton, WI (4). One of the was identical to T 811 except that a rigid platen loaded-cell
important features was preventing premature failure of the machine was used to measure the compression strength in-
test specimen at one of the loaded edges. A classical expedi- stead of the flexing-beam instrument. Subsequently, TAPPI
ent to eliminate edge failure is the use of a necked-down included both instruments in T 811 and withdrew T 823.
specimen shape such that the loaded edges are longer (e.g., Whitsitt and Schrampfer compared several test machines
3x) than the narrow necked-down portion where failure al- and concluded that the differences are not sigdicant. Fre-

Kroeschell is a consultant (retired vice president of technol- 'The Institute of Paper Chemistry is now the Institute of Paper Science
ogy, Stone Container), 272 Mooring Line Dr., Naples, FL and Technology, in Atlanta, GA.
33940, and Corrugated Containers Division Editor of Tappi
Journal.
January 1992 Tappi Journal 79
5
Y
quent instrument calibration is essential, regardless of the 1. The influence of pin adhesion values on ECT ( 7 7)
machine used.
3 Because the rigid-platen testing machine is easier to main-
tain in the parallelism required, it will produce more reliable

t
results day-in and dqy-out. It should be remembered that all
A
of the test data used to determine the strength levels in the
new Alternate Rules were determined on rigid-platen instru-
ments operated in accordance with T 823, now T 811. 100 -----------
Alternate procedures 8
+-
The use of alternate procedures has been the topic of wide s
discussion, and man,y literature references could be cited 90-
here. Major concern relates to the tedium in the edge rein-
forcement step. Australia has been strong in proposing that a
method adopted by FEFCO (the European federation of
corrugated box manufacturing associations) be an interna- I I I I 1 I I
P
tional standard (9). The U.S. concern with this widely used
method is that in this test, failures almost always occur at the PIN ADHESION, N/cm
loaded edges and the results are consistently lower than those
obtained by the T 811 procedure (about 12%on the average).
We contend that the FEFCO method does not measure the
intrinsic compression strength of the corrugated board. An-
other method has bee n developed by Schrampfer and Whitsitt
using a test fixture devised for the Japanese standard (JIS)
test method (IO). Again, the test shows mostly loaded edge
failure, but the results are equal to those obtainedwith T 811.
It is felt by many that use of the fixture is also tedious. Many in
slightly over 2 lb/in. or, for C-flute, only a little more than half
the industry worldwide feel that they have alternate proce- of the generally recognized minimum pin adhesion strength
dures that are simpler to use and highly correlated with T 811. specified for good operating practice. Thus, adhesion strength
The bottom line, however, is that the method prescribed in is of minimum importance in the edge crush test. This should
both the Rule 41 of the railroads or Item 222 of the trucks is not be surprising when it is realized that ECT is measured
TAPPI Official Test Method T 811. Other methods can be under the standard humidity of the laboratory. Only enough
used with confidence since many of these are highly corre- strength is needed to hold the component raw materials to-
lated with T 811, but for referee testing T 811 must be used. gether.
Glue skips are another factor. What should happen if a dry
The importance o:fconditioning line from one of the slots of the single-facer applicator roll
should occur in the test area? Eriksson reports a modest
Edgewise compression is extremely sensitive to the atmo- decline of ECT with increasing width of glue skip. A skip of 3
sphere in which the test is performed. Again, good correla- mm (1/8in.) can reduce the ECT by 5% (Fig. 2). Despite this
tions may be achievalde in the industrial atmosphere, but for minimal effect, it would seem that glue skips should be pre-
referee purposes the procedures prescribed in T 811 (refer- vented by proper setting of the clean-out prongs. Of course,
ence TAPPI T 402) for preconditioning, conditioning, and the fingerless single-facer makes this discussion academic
testing must be expli~~itly followed. unless glue skips are induced by factors other than the slotted
applicator roll.
Corrugatingvarialbles In unpublished work, McKee and his associates reported
that corrugated board with fractured flutes had a 15%reduc-
What are the variables that the boxmaker must control in
tion in ECT.
order to assure optimum compliance with the edge crush test?
What about crushing? Theoretically, crushing should also
Choice of raw materials is an obvious parameter and, indeed,
have a minimal effect on ECT. We should remember that the
as we will discuss later, it is the key variable. However, there
test heights in T 811 were chosen such that failure is in the
are other factors under direct control in the box plant such as
true compression mode. There should be no flexural failures
adhesion and absence of crushing that influence the test re-
in the test. This resulted in the recommendation of a flute
sults.
height of 38 mm for C-flute and 32 mm for B-flute. For
The most comprehensive review of the effect of boxmaking
example, ifwe were to crush the C-flute to the equivalentof B-
variables on the edge crush test was given by Eriksson (21).
flute, 28%,we could expect inducement of flexural failure and
He reported on the work of Langaard and Augustin, which
a significant reduction of ECT. At a series of seminars for
indicated that a pin adhesion value of only 4 N/cm was re-
boxmakers sponsored by The American Paper Institute and
quired to reach an eidge crush test independent of adhesion
the Fibre Boxhsociation in late 1984 and early 1985, Batelka
strength (Fig. 1). This equates to an adhesion strength of just
presented data which showed that this approximately 30%

80 January 1992 Tappi Journal

2. The relationship of ECT loss and box top-to-bottom compression 3. Loss of ECT, combined board thickness, and box top-to-bottom
strength to width of glue skips (72, 77) compression strength as a function of crushing (72, 75, 76)

>
- Top-to-bottom
compression
Ib facings)
-.---.
--n

E0 I compression
(26 Ibfacings)

Z I
W
0 -e ECT
- 0 - Recoveredcaliper
- Top-to-bottomcompression(ref. 15)
- - 0 -Top-to-bottom
- compression(ref. 16)

20' ' ' " " " ' ' 01 ' ' " ' ' ' " '
0 5 10 15 20 25 0 10 20 30 40 50
DRY LINE WIDTH, mm PERCENT CRUSHING ACTUALLY DONE

reduction in thickness resulted in a decrease of less than 15% search Institute of Canada, made an extensive study of the
in ECT (Fig. 3) (12). This was excessive crushing yet the relationship, in this instance using the short span compression
effect on ECT was modest. Where crushing was less severe test (TAPPI T 826) to characterize the edgewise compression
(less than lo%), ECT was virtually unchanged. Eriksson strength of the components. This has the advantage that the
reports that less than 40% reduction in thickness reduced scope of the test was not exceeded for any of the materials used.
ECT (by the FEFCO method) by less than 5% and that The study included 57 lots of commerciallymanufactured boards
complete crushing only lowered the ECT by 30%. of both B- and C-flute construction from three corrugated box
plants. Using the sum of the short span compression strength,
Linerboard and corrugating medium again adjusted by the take-up factor, correlation coefficients
with ECT ranged from 0.942 to 0.978. Seth concluded that, "In
Since all of these variables, adhesion, and combined board spite of the various factors that affect the strength of corrugated
thickness have only minimal effect on ECT, this property combined board and its measurement, this study has demon-
must be largely determined by the strength of the component strated that the edgewise compressive strength of combined
linerboard and corrugated medium; indeed, that is so. ECT is board is proportional to, and can be predicted with considerable
a material standard, but as we have said so very often in our accuracy from, the intrinsic edgewise compressive strengths
pronouncements regarding the new Alternate Rules, ECT is measured on corresponding component paperboards." In fact,
a material property much more closely related to box strength when we discussed the various test methods for ECT with the
than is bursting strength (Mullen), and it is a test that recog- Europeans, some Swedes asked, 'Why test ECT at all? Merely
nizes the contribution of the corrugated medium to box per- test the component materials and calculate ECT."
formance. If the variables under control by the boxmaker other than
What do we know about the relation of ECT to properties proper selection of the component paperboards have so little
of the linerboard and medium? Koning reported long ago that influence on ECT, and that is the criterion for strength in the
ECT was strongly related to the sum of the ring crush values new Alternate Rules, should we be concerned at all with crush-
of the component facings and the medium with allowances for ing, glue skips, and inadequate adhesion? The answer is "YES,"
the additional length of medium in a unit length of combined of course! While ECT is the single most important property of
board (the strength of the medium multiplied by its take-up corrugated board in determining box compression strength, it is
factor), with a constant multiplier to reflect the different unit not the only factor by any means. The second important prop-
lengths in the two tests (5).There is some deviation of this erty is the board's flexural rigidity, and this is determined by the
relationship in the extremes because the ring crush test was board's thickness (14).
being used to measure materials beyond its scope. However,
the relationship has good correlation and this has been con- The effect of crush on box compression
firmed by a number of published and many more unpublished
studies. Most recently Seth ( I @ , at the Pulp and Paper Re- In 1974, in a cooperative research project for the American

January 1992 Tappi Journal 81

 Paper Institute, Whitsitt and McKee, at the Institute of Pa- narrow glue skips was negligible on boxes made with heavy
Y
per Chemistry, subjected commercially manufactured boxes (229 g/mz)(47 lb/lOOO ftz)facings, on boxes made with 127 g/
8 to flat crushing loads 1,o determine the effect on box compres-
sion (25).Two sizes of boxes were fabricated from standard
linerboards and mediums of varying flat crush strength. Top-
m2(26lb/lOOO ft2 )facings, a glue skip 6 mm (0.24 in.) in width
caused an 18%reduction in compression strength.

load and end-load compression tests were made. Since vary- summary
ing flat crush boards were made and two crushing loads were
chosen, the crush varied from 30% to 93%of the boards’ flat The choice of ECT as the measure of strength in the new
crush strength. Under this range of loading the caliper was Alternate freight regulations is wise, for now we have a mate-
reduced by up to 1.16 mm (0.046 in.) or 29% of its original rial standard much more closely related than burst to con-
thickness. Top-to-bottom compression strength was decreased tainer performance. However, we must also be aware that
by 18-22%. These may seem to be modest reductions when some fabrication variables may not strongly influence ECT
related to the degree of crushing, but it should be noted that while they can have a very significant effect on performance.
these maximum reductions occurred even at 25% reduction in Ignoring these principles will produce box failures in ship-
caliper with loads only about 60% of the boards’ flat crush. ping, handling, and warehousing. Too many such failures
End-to-end loading w,asmore severely affected. Compression could lead to cancellation of the Alternate Rules. 0
strength losses were :is high as 41%.
Perhaps of even @*eatersignificance was the loss of com- Literature cited
pression strength of the uncrushed boxes. Compression
1. TAPPI T 811, “Edgewise Compressive Strength of Corrugated Fi-
strength was significantly lower than expected from theoreti- berboard (Short Column Test)’’, Atlanta, GA.
cal consideration of the ECT of the board. It was then noted 2. Kroeschell, W.O., Tappi J. 67(10): 56(1984).
that all boxes showed two relatively narrow crushed streaks 3. Kellicutt, K., Package Engineering 4(2): 60(1959).
running vertically on each box, presumably from improperly 4. McKee, R., Gander, J., and Wachuta, J., Paperboard Packaging
46(11): 70(1961).
adjusted pull-rolls on the folder-gluer. There was insufficient 5. Koning, J.,Tappi 47(3): 7(1964).
material to rerun the test. Only when the lowest caliper of 6. Koning, J., Tappi J. 71(10): 62(1988).
these two streaks was used in the box compression prediction 7. Schrampfer, K. and Whitsitt, W., “Combined Board Edge Crush
formula was there any agreement with the actual compres- (ECT) Technology”-Progress report to the American Paper Insti-
tute, The Institute of Paper Chemistry, Appleton, WI (Feb. 27,
sion measured. In other words, the maximum degree of crush- 1987).
ing determines the box’s ultimate compression strength. This 8. McLain, T. and Boltnoff, R., Tappi 65(3): 148(1982).
underscores the importance of eliminating crushing. Crush- 9. Stott, R., Tappi J. 71(1): 57(1988).
ing in a narrow streak. is just as damaging as an overall crush. 10. Schrampfer, K. and Whitsitt, W., Tappi J. 71(10): 65(1988).
11. Eriksson, L., “A review of the edge crush test,” Boxboard Contain-
Bick (16) has expressed these kinds of losses in terms of ers 86(8,9): 34,64(1979).
the cost of replacing the components with heavier materials to 12. Batelka, J., Proceedings of the Box Makers Seminars 1984-85, The
achieve the same compression strength had there been mini- American Paper Institute, New York, NY.
13. Seth, R., Tappi J. 68(3): 98(1985).
mal crushing. He shows an increase of fibrous raw material 14. McKee, R., Gander, J., and Wachuta, J., Paperboard Packaging
cost in excess of 20% if the board is crushed by just 0.5 mm 48(8): 149(1963).
(0.020 in.). He says, “...savings from the elimination of as little 15. Whitsitt, W. and McKee, R., “Study of the Relationship Between
as 0.002” of crush can result in savings approaching US$ Concora of Medium and Convertibilityand Box Performance,” Sum-
mary report to The American Paper Institute, The Institute of
100,000 per year.’’ The damaging effect of crushing will not Paper Chemistry, Appleton, WI (Dee. 27,1974).
make much of a mark on ECT and will therefore not affect the 16. Bick, G., Prime Technology, (10): 6(1989).
box’s acceptance by the freight regulations, but it can have a 17. Koning, J. and Moody, R., Tappi 52(10): 1910(1969).
significant effect on performance, as illustrated in Fig. 2.
We acknowledge, with thanks, permission of the American Paper Institute
There is compelling reason for the boxmaker to inspect every to cite sponsored research at the Institute of Paper Chemistry and permis-
place in his operation where crushing can occur and take sion to quote from one of its seminars. We also thank Messrs. G. Joseph Bick
corrective action when necessary, from the rider-roll on the and John W. Koning Jr. for their helpful suggestions in review of the
glue machine to the pull-rolls in the counter/ejector. manuscript.

Received for review July 29,1991.

The importance of adhesion on box performance
Accepted Sep. 24,1991.
Earlier we discussed the need for only minimum adhesion to
Keywords: Compression strength, corrugated boards, corrugated boxes,
the under standard humidity’ It
remembered that much research and practical experience
be __
crush resistance. edge crush tests, shipping
test methods, variabies.
- containers, stacking strength,

has shown that the stronger adhesion provided by water-

resistant bonding has,significantly improved the compression
strength and stacking performance of corrugated shipping
containers in the real environment.
Koning and Moody made an extensive study of the effect of
glue skips on the compression performance of corrugated
boxes (see Fig. 2) (27). They found that while the effect of

82 January 1992 Tappi Journal