| Material Information
||Comparative accuracy of certain open-side and open-end fertilizer sampling tubes
||Department of Soils mimeographed report 57-1 ; University of Florida
||7, 2 leaves : ; 28 cm.
||Volk, G. M ( Gaylord Monroe ), 1908-
Myers, J. Mostella ( Julian Mostella ), 1921-
University of Florida -- Dept. of Soils
University of Florida -- Agricultural Experiment Station
||University of Florida, Agricultural Experiment Station
||Place of Publication:
||July 2, 1956
||Fertilizer equipment -- Testing -- Florida ( lcsh )
Fertilizers -- Testing -- Florida ( lcsh )
||government publication (state, provincial, terriorial, dependent) ( marcgt )
non-fiction ( marcgt )
||Statement of Responsibility:
||by G.M. Volk and J.M. Myers.
||"July 2, 1956."
| Record Information
||University of Florida
||University of Florida
||All rights reserved by the source institution and holding location.
||oclc - 310116369
DEPARTMENT OF SOILS MIMEOGRAPHED REPORT 57-1
July 2, 1956
COMPARATIVE ACCURACY OF CERTAIN OPEN-SIDE AND OPEN-END
FERTILIZER SAMPLING TUBES
G. M. Volk and J. M. Myers
University of Florida
Agricultural Experiment Station
J. R, Beckenbach, Director
Comparative Accuracy of Certain Open-Side and Open-End
Fertilizer Sampling Tubes
G. M. Volk and J. M. Myers1
Investigations under the general Fertilizer Control Research Project
indicated that the official Florida sampling tool used in collection of
control samples might be causing a negative bias in nitrogen recovery and
a corresponding positive bias in certain other ingredients in incomplete
mixtures deriving nitrogen largely from pelleted materials. The Florida
tube is a simple slotted open-side, closed-end tube 3/4 inch outside
diameter with 30 inch penetration depth.
To examine this source of bias, various sampling tools were devised
and tested by sampling a 10-0-10 fertilizer made up from pelleted ammonium
nitrate, fine crystal muriate of potash and builder's sand. The fertilizer
was placed in a box 6 x 6 x 30 inches, over one end of which was clamped a
section of standard 4-ply fertilizer bag through which the various tools
could be inserted. The cover of the box was inset so that compaction
pressure could be applied for certain of the tests.
A minimum of eight and a maximum of 24 replicate cores were drawn in
various tests. Each core was individually put into solution, and the
solution analyzed for ammoniacal nitrogen as a measure of pelleted ammonium
nitrate and for chlorine as a measure of muriate of potash recovered for a
given core. The fertilizer was placed in the box in four different ways:
1 Dry mixture in loose condition.
2 Dry mixture in compact condition.
3 Moist mixture in loose condition,
4 Moist mixture in compact condition,
Soils Chemist and Associate Agricultural Engineer, respectively.
Moisture was supplied to 3 and h by adding water to the builder's
sand before mixing. 'his amount of moisture in the mixture was barely
perceptible to the touch, but was enough to develop some adhesion of
materials without making them sticky. Compacted mixtures removed after
completion of a test were very friable.
The sample tubes tested were based on two principles. They were the
open-side closed-end tubes represented by the official Florida tube, and
open-end cone-shaped tubes developed by the writers. Tests with various
tubes indicated four possible sources of bias in recovery of pelleted
I Disproportionate entry of fine materials as compared to pelleted
materials in a loose dry mixture entering an open-side tube.
II Segregation of materials around the perimeter of a hole forced
into mixtures by a solid penetrating point. The geometry of
particle size distribution would tend to give a disproportionately
large amount of fines at the perimeter as a result of forcing
back of larger particles.
III Formation of a compression and shear zone ahead of a tube. The
process tends to develop segregation to the extent that the
compression cone most resistant to shearing differs from the
composition of the mixture.
IV The relation of curvature of the circumference of an open-end
tube to the diameter of the larger pellets in the mixture. As
greater curvature is introduced by decreasing diameter, the
tendency for larger pellets to move out instead of in becomes
The order of development of tubes resulted from the order of recog-
nition of possible sources of bias. This explains the non-orthogonal
nature of the tests. Statistical comparisons were made by using the
average standard deviation of 4.34, and making comparisons individually
to values obtained by the Florida tube for the comparable moisture and
compaction status, based on the standard error for the least number of
replicates in the pair of values being compared.
Data are summarized in Table 1. The individual standard deviations
are included because they have certain indicative value over and above
their use in determining significance.
Descriptive names and letters for the various tubes have the
Official Florida This is the tube exactly as used for the collection
of official samples.
Florida F The official tube with a I inch fin welded the length of the
tube on one side of the slot so as to give a deeper "bite" as the tube
is turned in filling.
Florida PDF The official tube with longer and sharper point and fitted
with a finned sleeve thus making a double wall tube that would permit
closure during penetration and withdrawal. The long point was intended
to reduce the compression cone, the fin to give deeper "bite", and the
closure system to prevent rakeoff of surface pellets during withdrawal.
VM This refers to the open-end cone principle embodied in a series of
tubes. The tube was constructed of stainless steel sheet welded into a
cone 1 1/8 inch diameter at the large end and decreasing in diameter at
- inch per foot toward the cutting or insertion end. The associated
figure in inches refers to the insertion depth measured from the large
end, after allowing 2- inches projection for a handle.
VM 20" S Cut at 45 slant with inside diameter of 0.60" at cutting end.
VM 15" Cut vertical. 0.75 inside diameter.
VM 15"SC Cut at 780 slant. 0.75" inside diameter except the heel of
slant depressed to reduce diameter to 0.63" and provide additional con-
striction at the opening into the tube. It was intended that the long
point of the slant would separate the mixture without bias before a
pressure cone developed, and that the depressed heel would allow rapid
release of pressure within the tube end to aid unbiased entry.
VM 15"C Vertical cut but constricted perimeter to give 0.65" inside
diameter of cutting edge.
VM 10" Cut vertical. 0.85 inside diameter.
VM 10"C Constricted like VM 15"C to give 0.77" inside diameter.
VM 10"B Cut vertical with ring inserted to give blunt with 0.69" inside
diameter and marked constriction.
The Florida type tubes were inserted with gradual even pressure in
inverted position, and then turned to place the slots upright for with-
drawal. The VM type tubes did not obtain the amount of sample to be
expected when used with gradual pressure. Driving these tubes with sharp
blows with a light ballpeen hammer proved to be the best procedure. This
took advantage of the inertia of the material to move it into the tube
with least bias.
Data in Table 1, fourth column, show first that the Florida tube
introduced a definite negative bias in recovery of pelleted ammonium
nitrate in dry mixtures. Modifying the Florida tube gave no improvement.
The best recovery with this tube was from moist loose mixtures where dry
segregation of fines, of Type I bias did not occur, and where there was
not sufficient resistance to penetration to develop much segretation of
Type II. VM tubes did not improve on accuracy with this mixture. This
agrees with results from complete mixtures sampled under the general
Fertilizer Control Research Project.
The VM tubes improved accuracy in practically all instances where
the Florida tube was grossly in error. The 15 inch tube gave significantly
increased recovery of ammonium nitrate in all instances, except when a tube
with constriction around the entire perimeter was used; or with the loose
moist mixture where the Florida tube itself gave little bias. The improved
accuracy amounted to reduction of the bias by 50 to 60 percent.. Apparently
the slant cut and partial constriction of VM15"SC is not necessary. A tube
of greater diameter would reduce bias still more, and such a tube may be
practical provided damage to the bag is not a limiting factor. Constriction
around the complete circumference of the tube appears to be undesirable,
probably because of greater development of a compression cone.
Figure 1 shows the type of positive bias usually occurring for muriate
of potash when a negative bias of pelleted ammonium nitrate exists. This
is a summary of all samples analyzed for Table 1. If the pelleted nitrogen
does not enter the tube in proper proportion, then it must be substituted
with fines to take its place. The degree to which this negative correlation
occurred is shown by the graph, The fact that the regression lime passes
through the 100 percent recovery intersection indicates the accuracy of the
The Florida tube collects from 100 to 150 grams of sample per probe,
while the VM15" tube collects 100 to 130 grams per probe depending on
dryness and compaction of mixtures. The best construction for the VM15"
tube and protective driving head to withstand field usage is yet to be
decided. The head used for the laboratory tests was a 2 x 2 inch piece
of oak flooring to the reverse side of which was nailed a rubber stopper
of proper size to center it and hold it to the tube end. There was no
evidence of clinging or compaction in the cone shaped tube. A simple
cylindrical tube was not considered because experience with soil tubes
indicates that jamming sufficient to increase the compression cone would
result even if constriction at the cutting edge were included.
Depth of penetration of a tube is important in obtaining a long cross
section of material, and to properly sample bagged mixtures without con-
sistently introducing the bias resulting from segregation within bags.
The official tool with 30 inch penetration is intended to be used diagonally
from corner to corner of a sack lying on its side. The principle is sound,
but in practice the most biased area probably lies next to the sack wall
and may not be properly sampled because of danger of double puncture of
the bag at extreme penetration. A substitute would be 2 penetration,
assuming sacks are sampled from either end with equal frequency. This
would eliminate one serious objection to the VM sampler the large
diameter and resulting bag damage necessary to give penetration depth
equal to the Florida tube, yet retain the cone shape essential to free entry
of the sample. A sampler of 15 inches penetration depth appears to be
The primary contribution of these tests is the principle of the cone-
shaped open-end tube, and the use of sharp, light driving blows rather
than pressure as the method of insertion. The value of moisture or a
similar factor to increase cohesion is indicated. It is suggested that
further testing must now be done under actual field conditions. The
most promising tube to compare to the official Florida tube would be VM1N"
without slant or constriction.
In using this tube it is necessary first to make a hole in the bag
with a pointed probe so that the point of the tube can be inserted without
carrying a wedge of paper and restricting freedom of entry of the sample.
The weight of the driving hammer will be determined by the weight of the
tube plus head, and by resistance to penetration encountered in routine
use. Its weight should be such that a sharp blow will move the tube
between one-half and one inch.
250 copies Soils Department 7/5/56
Table 1. Percentage Recovery of Nitrogen from 10-0-10
Made up From Pelleted Ammonium Nitrate
Fine Crystal Muriate of Potash and Builder's Sand.
S: : : Statistical :
: : No, of : Standard : Percentage : Significance :
: Tube : Samples : Deviation : Recovery of : over Off. :
: Type : Drawn : (Av. 4.34) : Nitrogen : Florida Tube :
: Dry Mixture in Loose Condition
: Off.Florida : 16 : 6,02 : 908 :
: VM 20" S : 8 : 3.36 : 93.4 : No
: VM 15" : 8 : 3.29 : 96.7 : Yes (.01)
: VM 15" SC : 16 : 6.45 : 94.6 : Yes (.01)
: VM 15" C 8 : 2.97 : 91.8 : No
: Dry Iisture in Compact Condition
:Off.Florida : 24 : 4.40 : 90.2
: Florida F : 8 : 3.21 : 90.3 No
: Florida PDF : 8 : 5.7 : 87.2 :No
:VM 15" 9 : 3.60 : 9.8 : Yes (.05)
SVM 15" SC 16 : 7.71 : 95.8 Yes (.01)
: VM 15" C 8 2.61 : 90.3 :No
: VM 10" 8 : 6.83 : 100.9 : Yes (.01)
: Moist Mixture (0.44% water) in Loose Condition
: Off.Florida : 24 : 5.78 : 97.2* :
:Florida F : 8 : 3.6 : 95. No
SV 20" S : 16 : 2.81 : 95.6 No
:V 15" 8 : 2.59 : 97.1 : No
: Moist Mixture (0.h44 water) in Compact Condition
: Off.Florida : 18 : 5.38 : 93.2- :
: V 15" : 16 : 2.79 : 97.7 : Yes (.01)
: VM 15" SC : 22 : 4.79 : 97.6 : Yes (.01)
:VM 10" : 8 : 4.20 : 9.1 : No
:VM O" B : 8 : 3.2 : 96.5 : No
* Official Florida tube gave significantly greater recovery (.01)
from moist mixturesthan from dry mixtures under similar compaction.
Percentage Potash Recovery
p. 0- 0
0 :z 0
o* a 0
I .. -. 1.. -~ --- .