Slab On Grade
Slab On Grade
Concrete Slabs-on-Grade:
Warehouses II – Slab Design
by
Course Outline:
This course is a continuation of the first warehouse course, which is Concrete Slabs-on-
Grade: Warehouses I – Background & Loading
The following figures are added for convenient reference to the first warehouse course.
https://commons.wikimedia.org/wiki/File:CEDIS_Soriana.jpg
Organizacion Soriana / CC BY-SA (https://creativecommons.org/licenses/by-sa/4.0)
Fig. 25 - Stacker lift making a placement/retrieval of a pallet
Here begins the continuation from the first warehouse course, which is Concrete Slabs-
on-Grade: Warehouses I – Background & Loading.
For a quick recap of some of the design options available are added here for convenient
reference to the first course.
The list below is based on a list noted in Appendix D of the FEMA 460 document,
Seismic Considerations for Steel Storage Racks located in Areas Accessible to the
Public. It is as good a list as can be found on the subject of slab design. Availability of
publications noted may vary.
“The information presented below was issued by the City of Los Angeles Department of
Building and Safety as Information Bulletin/Public-Building Code, Reference L.A.M.C.
91. 1806,
Document P/BC 2002 to be effective May 10, 2004…
“PURPOSE: This Information Bulletin establishes a list of acceptable analysis methods
for slabs-on-grade
(“SOG”) as foundations.
“ACCEPTABLE DESIGN METHODS: The following methods of design and analysis for
SOGs are acceptable:”
• Packard, Robert G., “Slab Thickness Design for Industrial Concrete Floors on Grade,”
IS195.01D, Portland Cement Association, Skokie, Illinois, 1976.
• Departments of the Army and Air Force, Concrete Floor Slabs on Grade Subjected to
Heavy
Loads,” ARMY TM 5-809-12, Air Force AFM 88-3, Chapter 15, 1987.
• Empirical Method - Method of analysis based on studies that compare the load test
results to
computer analysis. (Shentu, L., Jiang, D., Hsu, T. (1997). “Load Carrying Capacity for
Concrete Slabs on Grade.” Journal of Structural Engineering, ASCE, January 1997, pp
99-103.)”
The author would add to this list the following very comprehensive and helpful
publication from overseas, which is based on a yield line-type method, but might be
limited to reinforced slabs upon review due to the ultimate limit state thereby required
for use:
• Technical Report 34 (TR 34), “Concrete Industrial Ground Floors - A Guide to Design
and Construction“, The Concrete Society, 2003.
Also, the following is a document based on the PCA method that is one of three
methods noted in ACI-360. It is one that comes highly recommended from the author,
and is a document that has been developed over many years. The fourth edition is
recommended for today’s slabs. Earlier versions may be dated. Combined with ACI
360, these are the definitive warehouse design publications.
• Tarr, Scott M., and Farny, James A.; Concrete floors on ground. Portland Cement
Association. Skokie, Ill., 2008.
And lastly, a direct analytical approach could be used as well, especially for very
unusual or unique loads. The author would round this list out with:
In Fig. 30 from ACI 360, the effects of subbase thickness on the modulus of subgrade
reaction are shown. In design, an envelope of soil stiffness conditions with upper and
lower bounds can be studied.
In designing for slab thickness, the stiffness of the soil below the slab is an important
factor in analytical or design tools. If the soil were as stiff as concrete, for example, and
we assumed that the slab was free of voids (such as those caused by curling), loads
would more or less be transferred in bearing to the soil below. Decreasing soil stiffness
by degrees changes the behavior from bearing to that which depends on the relation
between concrete section properties and the soil stiffness. In the 1920’s, Westergaard
developed methods for determining stresses in concrete road slabs. In many of the
available analytical or approximate methods used to determine stresses in loaded
building slabs, some form of Westergaard’s work is generally used. The deformed
shape of the slab, compatible shape of the soil, and the relative stiffnesses of the slab
and soil are considered (Westergaard).
Maximum tensile stress in the top of the slab, as the slab acts like a short cantilever
supported by soil, is given by:
0.6
3 𝑃𝑃 𝑎𝑎√2
𝑓𝑓𝑡𝑡 = 2 �1 − � � �
ℎ 𝐿𝐿
𝑃𝑃 = 𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿, 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝
𝑎𝑎 = 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑜𝑜𝑜𝑜 𝑡𝑡ℎ𝑒𝑒 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 (𝑜𝑜𝑜𝑜 𝑜𝑜𝑜𝑜𝑜𝑜 ℎ𝑎𝑎𝑎𝑎𝑎𝑎 𝑡𝑡ℎ𝑒𝑒 𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤ℎ 𝑜𝑜𝑜𝑜 𝑎𝑎 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝)
𝐸𝐸 = 𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 𝑜𝑜𝑜𝑜 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑜𝑜𝑜𝑜 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐, 𝑝𝑝𝑝𝑝𝑝𝑝
ℎ = 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑡𝑡ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖, 𝑖𝑖𝑖𝑖
𝑓𝑓𝑡𝑡 = 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑒𝑒 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑖𝑖𝑖𝑖 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐, 𝑝𝑝𝑝𝑝𝑝𝑝
4 𝐸𝐸ℎ3
𝐿𝐿 = �
12 (1 − 𝜇𝜇 2 )𝑘𝑘
𝐿𝐿 = 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑜𝑜𝑜𝑜 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠, 𝑖𝑖𝑖𝑖
𝑘𝑘 = 𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 𝑜𝑜𝑜𝑜 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟, 𝑝𝑝𝑝𝑝𝑝𝑝
𝜇𝜇 = 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑛𝑛′ 𝑠𝑠 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑓𝑓𝑓𝑓𝑓𝑓 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐, 𝑎𝑎𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝. 0.15
Case 2: Load located far away from the edges of the slab:
Maximum tensile stress in the bottom of the slab is given by:
𝑃𝑃
𝑓𝑓𝑏𝑏 = 0.316 2 �𝑙𝑙𝑙𝑙𝑙𝑙[ℎ3 ] − 4 log ��1.6 𝑎𝑎2 + ℎ2 − 0.675 ℎ� − log(𝑘𝑘) + 6.48 �
ℎ
Case 3: Load at the edge of the slab, but far away from the corners:
Maximum tensile stress in the bottom of the slab is given by:
𝑃𝑃
𝑓𝑓𝑏𝑏 = 0.572 2 �𝑙𝑙𝑙𝑙𝑙𝑙[ℎ3 ] − 4 log ��1.6 𝑎𝑎2 + ℎ2 − 0.675 ℎ� − log(𝑘𝑘) + 5.77 �
ℎ
𝑓𝑓
𝐹𝐹𝑏𝑏 = 𝐹𝐹𝐹𝐹𝑟𝑟 > 𝑓𝑓𝑏𝑏 𝑜𝑜𝑜𝑜 𝑓𝑓𝑡𝑡
ACI 318 uses a value of 7.5 �𝑓𝑓𝑐𝑐′ for rupture strength, where 𝑓𝑓𝑐𝑐′ is the compressive
strength of the concrete in psi. To calibrate project-specific concrete, testing can be
done on the concrete to estimate the bending strength, such as ASTM C78. The author
has used a higher value of rupture modulus based on research, in the range of 9�𝑓𝑓𝑐𝑐′ at
28 days, and this value has been confirmed by ASTM C78 testing. It should be noted
that a factor of safety will be applied to obtain an allowable stress. This is an
engineering judgment call that needs to be reviewed by the Engineer of Record. Factors
of safety will depend on a number of things, including, but those used are typically in the
1.5 to 2.1 range.
The use of any of these methods should entail a closer review of other behavior, such
as curling, cracking, slab deflections under load, and load transfer at joints. Also,
adequate ductility of the concrete will be required for plastic or ultimate limit state
methods. If plastic or ultimate limit state methods are used, it is recommended that a
means of providing ductility and post-crack residual strength be used. This would
include steel bar reinforcing, heavier steel mat reinforcing, steel fibers, or synthetic
macrofibers. Application of elastic methods by PCA are described first, and then some
of these plastic or ultimate limit state methods are described below.
To be fair, most methods assume no curling stresses, and, in fact, a slab that maintains
contact with the ground. Some adjustment should be made for this, as a curled slab will
behave differently where it has lifted off of the base soil. Conceivably, finite element
analysis could include geometry and/or compression-only spring behavior, along with
curled edges with gaps to model curling. Or, these additional behaviors could be
factored in directly with some estimated stresses. In a design example for slabs with
steel fibers, ACI 360 uses a value of 200 psi flexural tensile stress for combined
shrinkage and curling stresses in the analysis.
The PCA method is tailored for warehouse-type buildings, with various approaches to
wheel and rack post loading, as well as finding a slab thickness for a given aisle width
with uniform loading on either side. This design method is the basis of PCA’s historical
document, Concrete Floors on Ground. Obtaining the latest edition is recommended.
Starting with the 4th edition, design charts began to include joint load transfer conditions
explicitly. Earlier versions may be dated. It is not useful to repeat the charts here without
the accompanying extensive and nuanced background, but the publication is readily
available via pdf for a very reasonable price ($45 for the pdf version as of October,
2020).
In Chapter 5 of PCA’s Concrete Floors on Ground, there are tables and charts that are
based on common warehouse rack configurations, uniform floor load arrangements,
and vehicle-types. The publication is extremely useful for typical warehouse slab
design. Design methods are available for the following loading.
1. Vehicle Loading
0.90 7
0.95 2
1.00 0 (emprical value, not from equation above)
The factor of safety for a fatigue failure is the inverse of the stress ratio. PCA
uses the term SF for factor of safety for fatigue.
So, SF = 1/SR
SF = 2.2 amounts to unlimited traffic. Per PCA, this value might be used for
heavily trafficked areas. A lower SF of 1.4 to 2.0 might be used in lower traffic
areas.
Joint load transfer is touched on in Slabs-on-Grade: From the Ground Up, and
also discussed later in this course. Vertical load transfer is any means keeping
opposing sides of the slab at a joint from moving separately as a wheel load
crosses the joint (see Fig. 45). If the transfer is there, it drastically reduces stress
at the joint. There are two main choices for providing a means of vertical load
transfer:
i) Aggregate interlock (friction between the slabs on either side of a
sawcut-induced crack). The effectiveness of aggregate interlock
depends on the joint size. (Fig. 34)
ii) Dowels providing full load transfer at a joint. (Fig. 45)
If aggregate interlock is used, the following table is used to obtain a joint factor
(JF) as long as the joint spacing is reasonable (based in judgement) to factor in
effectiveness. This joint factor is used similar to a factor of safety when
determining allowable stresses.
c. PCA Charts for vehicles – steps explained further in the PCA publication.
Again, the charts aren’t presented here.
i) From facility data, equipment info needed:
1. Axle load
2. Wheel spacing
3. # of wheels on an axle
4. Tire type, width, contact length
ii) Data for slab/soil needed:
1. Subgrade modulus
2. Concrete flexural strength
3. Ultimate shrinkage (at infinite time)
4. Joint spacing
iii) Figure out working stress using SF and JF
𝑀𝑀𝑀𝑀
1. 𝑊𝑊𝑊𝑊 = 𝑆𝑆𝑆𝑆 𝑥𝑥 𝐽𝐽𝐽𝐽 (MR = Flexural strength, use MR = 9 �𝑓𝑓𝑐𝑐′ )
2. Note that at this point, the slab designer could factor in curling
stresses if desired by reducing MR to get a reserve strength
value.
iv) Then find working stress per 1000# of axle load
1. Value = WS/(Axle load,kips)
v) Enter charts using based on type of joint assumed and find a slab
thickness
2. Post Loading
a. Design for post loading is based a single controlling design event (maximum
post loads expected for a given rack post layout) rather than fatigue loading.
b. PCA Charts for vehicles – safety factors and steps explained further in the
PCA publication. A quick reminder that the charts aren’t presented here, and
that the publication itself is necessary to complete the slab design.
i) From facility data, equipment info needed:
1. Maximum post loads
2. Post spacings, x and y in plan
3. Base plate area
4. Tire type, width, contact length
ii) Data for slab/soil needed:
1. Subgrade modulus
2. Concrete flexural strength
3. Ultimate shrinkage (at infinite time)
iii) Review notes in the publication on joint spacing and how slab edge
stress may be reduced for good load transfer at the joint.
iv) Note in particular the wider range of factors of safety involved
depending on the application, as a racking collapse would be much
more dangerous than fatigue failure of joints.
v) Figure out working stress using SF and JF
𝑀𝑀𝑀𝑀
1. 𝑊𝑊𝑊𝑊 = 𝑆𝑆𝑆𝑆 𝑥𝑥 𝐽𝐽𝐽𝐽 (MR = Flexural strength)
2. Note that at this point, the slab designer could factor in curling
stresses if desired by reducing MR to get a reserve strength
value.
vi) Then find working stress per 1000# of post load
1. Value = WS/(Post load,kips)
vii) Enter charts using based on type of joint assumed and find a slab
thickness.
3. Distributed Loading
a. This section of the PCA publication addresses distributed loads on slabs,
including those occurring next to open travel aisles free of loads. The charts
and tables aren’t presented here due to the amount of nuanced background
needed to use them.
The charts in the PCA publication address an assumed wheel load for a given number
of repetitions to achieve a given factor of safety with a given wheel spacing and contact
area. Thus, the use of the tables necessitates the use of a single forklift load that
represents all traffic. A suggested alternate method is described below.
If desired, the author proposes that an approximate study could be done directly if the
wheel loading varies. Note that this is not part of the information presented by PCA. For
a wide variety of loadings on a given slab, it may be necessary to divide the loads into
categories by magnitude of loading and number of cycles, and the contributions of each
category contribute to the total accumulation of damage. For example, if the amount of
accumulated damage to failure is 1.0, and we divide loads into several categories, with
various stress ratios for each category, we can examine the damage as follows:
i = number of the category (1, 2, 3, … m)
ni = number of cycles in each category
Ni = the number of loads to produce a flexural fatigue cracking failure for category i
SRi = Stress ratio for category i
Steps:
Step 1) Find out what the facility intends to use for equipment, what loads might be
imposed by them, and how many cycles of load there might be for each loading. Here it
might be a collection of good guesses, unless there is data to draw from.
Step 2) Divide the loads into a number of categories with a relevant SRi for loading
significacnt enough to load to failure (SR > 0.45). The loading under study might be the
stress in the slab as the wheels cross joints, with the aid of Westergaard equations. It is
recommended that the PCA method be studied for nuances in design choices.
Step 3) Use the equations or tables to get the values Ni. Apply a factor of safety SF and
joint factor, JF in the manner described for a single type of vehicle by PCA.
Step 4) Figure out whether fatigue limits are met:
𝑛𝑛𝑖𝑖
If ∑𝑚𝑚
𝑖𝑖=1 𝑥𝑥 𝐹𝐹𝐹𝐹 𝑥𝑥 𝐽𝐽𝐽𝐽 < 1, fatigue limits are OK. If the value of the expression on the left
𝑁𝑁𝑖𝑖
is more than 1, a thicker slab may be required.
The generally accepted model for the failure response of a concrete slab-on-grade to a
point load applied via a round plate is shown in Fig. 31 (Several Sources: TR34, Walker
and Holland (2001), Holland, World of Concrete). The deformation of the slab is circular
dish shape with characteristic defining radii that will be described below. Progressive
failure for gradually increasing load begins when the slab first sags in the middle, with
concave shape up, developing positive moments (slab in tension on the bottom and
compression on top) along radial lines. Then the slab bends down, or hogs, in
negative moment (slab in compression on the bottom and tension on top), at a
characteristic radius from the load center. The negative moment increases until failure
occurs when the cracking moment is reached on top of the slab.
Side note: For serviceability crack concerns for loads less than ultimate loads but
greater than that required to develop initial cracking, if bottom of slab cracking controls
the design, the post-initial-crack behavior may be an acceptable result. (Holland and
Walker, 2001).
In TR34, the yield line failures that take place along positive moment regions are
assumed to occur in a ductile manner due to concrete containing steel or synthetic
fibers or steel reinforcing. Minimum amounts of reinforcing are recommended. The
literauture mainly focuses on concrete slabs with steel fibers, but the use of the
equations could be extrapolated to slabs with steel reinforcing.
The equations for rack post loading are as follows:
4 𝐸𝐸𝑐𝑐 ℎ3
𝐿𝐿 = �
12 (1 − 𝜇𝜇 2 )𝑘𝑘𝑠𝑠
Where,
𝐿𝐿 = 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑜𝑜𝑜𝑜 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠, 𝑖𝑖𝑖𝑖
ℎ = 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑡𝑡ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖, 𝑖𝑖𝑖𝑖
𝜇𝜇 = 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑛𝑛′ 𝑠𝑠 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑓𝑓𝑓𝑓𝑓𝑓 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐
Case I: Internal load – center of load is more than (L+a) from a free edge or joint
a/L = 0, 𝑃𝑃𝑢𝑢 = 2𝜋𝜋�𝑀𝑀𝑝𝑝 + 𝑀𝑀𝑛𝑛 �
𝑎𝑎
a/L > 0.2, 𝑃𝑃𝑢𝑢 = 4𝜋𝜋�𝑀𝑀𝑝𝑝 + 𝑀𝑀𝑛𝑛 �/ �1 − 3𝐿𝐿�
Use linear interpolation for a/L between 0 and 0.2.
Case II: Edge load - center of load is located on an edge more than (L+a) from a free
corner or the intersection of two joints
a/L = 0, 𝑃𝑃𝑢𝑢 = �𝜋𝜋(𝑀𝑀𝑝𝑝 + 𝑀𝑀𝑛𝑛 )/2� + 2 𝑀𝑀𝑛𝑛
2𝑎𝑎
a/L > 0.2, 𝑃𝑃𝑢𝑢 = ��𝜋𝜋(𝑀𝑀𝑝𝑝 + 𝑀𝑀𝑛𝑛 )/2� + 4 𝑀𝑀𝑛𝑛 � / �1 − 3𝐿𝐿
�
Use linear interpolation for a/L between 0 and 0.2.
Case III: Corner load - center of load is located at distance a from the two edges
forming a corner
a/L = 0, 𝑃𝑃𝑢𝑢 = 2𝜋𝜋𝑀𝑀𝑛𝑛
𝑎𝑎
a/L > 0.2, 𝑃𝑃𝑢𝑢 = 4 𝑀𝑀𝑛𝑛 / �1 − 𝐿𝐿 �
Use linear interpolation for a/L between 0 and 0.2.
For determining limits on proximity of adjacent loads, per TR3, “The influence of an
additional load… at a distance x from…[the load center] is as follows:
If x < L, “the positive bending moment at …[the load center] will increase”
If L < x < 3L, “the positive bending moment at …[the load center] will decrease, but by a
relatively small amount.”
If x > 3L, “the additional load will have negligible influence on the bending moment at
…[the load center].”
Cases for multiple wheel loads, line loads, and uniform loads are also presented in the
TR34 document, in the form of direct equations, making it very suitable for use in a
spreadsheet.
of safety, the failure loading predicted by the model is substantially greater than that
obtained by other methods of analysis.
As a caution in using this approach, the designing engineer will need to consider the
things which do not appear to be addressed in this research, including curling and
shrinkage stresses, and joint proximity and construction. Also, if applying this method to
the analysis of the slab due to seismic forces from a rack post, the uplift values on a
cracked slab during repeated cycles of alternating upward and downward forces due to
overturning need to be considered.
The use of the method developed by Shentu, et al, is listed as an acceptable method of
slab design in FEMA 460. An excellent summary of the use of this method is also found
in “Load Carrying Capacity, Concrete Slabs-On-Grade Subject to Concentrated Loads”,
by Azzi and Laird. They further recommend a reduction factor to account for slightly
different theoretical results for thicker slabs. The resulting equations are:
𝑘𝑘𝑠𝑠 𝑅𝑅1
𝑃𝑃𝑢𝑢 = 1.72 𝑥𝑥 � 𝑥𝑥 104 + 3.60� 𝑓𝑓1′ 𝑑𝑑 2
𝐸𝐸𝑐𝑐
𝑃𝑃𝑢𝑢
𝑃𝑃𝑎𝑎 =
𝐹𝐹𝐹𝐹
𝑃𝑃𝑢𝑢 = 𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑜𝑜𝑜𝑜 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠, 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝
𝑃𝑃𝑎𝑎 = 𝑎𝑎𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑜𝑜𝑜𝑜 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠, 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝
𝐹𝐹𝐹𝐹 = 𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 𝑜𝑜𝑜𝑜 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠
𝑅𝑅1 = 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝑜𝑜𝑜𝑜 𝑡𝑡ℎ𝑒𝑒 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 𝑜𝑜𝑜𝑜 𝑜𝑜𝑜𝑜𝑜𝑜 ℎ𝑎𝑎𝑎𝑎𝑎𝑎 𝑡𝑡ℎ𝑒𝑒 𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤ℎ 𝑜𝑜𝑜𝑜 𝑎𝑎 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝
𝐸𝐸𝑐𝑐 = 𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 𝑜𝑜𝑜𝑜 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑜𝑜𝑜𝑜 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐, 𝑝𝑝𝑝𝑝𝑝𝑝
𝑘𝑘𝑠𝑠 = 𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚 𝑜𝑜𝑜𝑜 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟, 𝑝𝑝𝑝𝑝𝑝𝑝
𝑓𝑓1′ = 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐, 𝑝𝑝𝑝𝑝𝑝𝑝
𝑑𝑑 = 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑡𝑡ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖, 𝑖𝑖𝑖𝑖
Where the load reduction factor, 𝛽𝛽, introduced by Azzi and Laird is given by
𝛽𝛽 = 1.0 for slabs < 7” thickness, and 𝛽𝛽 = 0.85 for slabs ≥ 7” thickness.
Joint Layout
Control joints, sawcutting, construction joints and isolation joints for unreinforced slabs
in typical buildings was described in detail in Slabs-on-Grade: From the Ground Up.
Fig. 32 - Sawcutting new concrete, and joint with crack induced by sawcut (MnDOT)
The same concepts apply to warehouse slabs. Control joint layout for warehouses
depend on a number of parameters. Shrinkage characteristics of the slab will need to be
understood. Preventing random cracks due to shrinkage restraint will be a goal, and
how much joints will open up due to shrinkage will also factor in. Obviously, the larger
the joint spacing, the larger the joint width can be expected with shrinkage. Expectations
of joint aggregate interlock performance will also need consideration, as was discussed
in Slabs-on-Grade: From the Ground Up, and further discussed below. Joints should be
kept perpendicular to traffic, with parallel joints free of main travel aisles. Luckily, there
is guidance from the industry. ACI recommends the following joint spacing for
unreinforced slabs (see Fig. 33), from PCA’s Concrete Floors on Ground. Note that the
upper and lower bounds are 24 and 36 times the slab thickness, respectively:
Walker and Holland (1999) performed finite analysis on modeled slabs and found that
slab thickness and joint spacing can affect curling stresses dramatically. They state that
15 ft joint spacing is significantly better for curling stresses than 30 ft joint spacing.
Additionally, an increase the modulus of subgrade can increase the curling stress,
especially for shorter joint spacings. So, efforts to stiffen the subgrade may not be
advantageous beyond a certain point. If we pick two slab thicknesses, 6” and 8”, results
were as follows:
a. 6” slab – 18 ft joint spacing, curling stress = 303 psi
b. 6” slab – 15 ft joint spacing, curling stress = 217 psi
c. 8” slab – 24 ft joint spacing, curling stress = 350 psi
d. 8” slab – 15 ft joint spacing, curling stress = 143 psi
The lower range of modulus of rupture for bending in concrete used in practice is 410
psi. If the curling stress is high, it reduces the reserve strength near the joint to
resist wheel loads. Walker and Holland also found that higher strength concrete results
in higher curling stresses due to the increase in concrete shrinkage and stiffness, both
resulting in greater curling. The lesson here is keep joints to a maximum of 15 feet if
possible, and use concrete in the range of 3,000 to 3,500 psi.
Again, the two types of joints that are used in traffic areas of slabs on grade are control
joints and construction joints. We will look at both of these separately, but first we can
explore the options, beginning with a discussion on aggregate interlock.
Load Transfer
One might consider aggregate interlock a reliable method for areas with lower loading
or minimal traffic volumes. However, rack layouts and facility usage may be revised in
the future. That being said, it is worthwhile to explore the limits of aggregate interlock for
any slab being considered. Colley and Humphrey performed research for PCA on slabs
to determine the effectiveness of aggregate interlock (Colley, B.E. and Humphrey, H.A.,
1967, “Aggregate Interlock at Joints in Concrete Pavements,” Development Department
Bulletin D124, Portland Cement Association, Skokie, IL.). They loaded plates near the
edges of a slab joints to see how well the load would be passed to the far side of the
joint. They defined “effectiveness”, E, as
2 𝑑𝑑𝑗𝑗′
𝐸𝐸 (%) = 𝑥𝑥 100
𝑑𝑑𝑗𝑗 + 𝑑𝑑𝑗𝑗′
where 𝑑𝑑𝑗𝑗 is the deflection of the slab at the near side of the joint 𝑑𝑑𝑗𝑗′ is the deflection of
the slab at the far side of the joint. An example curve of the results is shown in Fig. 35.
Colley and Humphrey found that the value of E depends on a number of things. They
summarized this in an empirical expression as noted below, where EI is the endurance
index, which is “obtained by dividing the area under the curve of effectiveness vs cycles
by the area that would be developed if the joint retained an effectiveness of 100 percent
throughout one million load applications.”
ℎ
𝐸𝐸𝐸𝐸 (%) = 230 √𝑘𝑘
𝑃𝑃 𝑤𝑤
Where:
h = depth of roughened surface, in
P = wheel load, #
w = joint opening, in
k = foundation modulus, pci
(per Colley and Humphrey)
Some specific data of note in the report was as follows. For a 7” slab with a 6” gravel
base to stay at 50% effectiveness at one million cycles, the joint opening needed to be
about .025”. For a 9” slab with a 6” gravel base to stay at 50% effectiveness at one
million cycles, the joint opening needed to be about .035”. The effectiveness of
aggregate interlock drops steeply with increasing joint size.
So, again, unless joint size or number of load cycles can both be practically controlled
or are naturally limited, caution should be exercised in applying this method of joint load
transfer. Per ACI 360: “If the designer cannot be sure of positive long-term shear
transfer at the joints through aggregate interlock, then positive load-transfer devices
should be used at all joints subjected to wheeled traffic.”
The function of dowels is to allow for joints to continue to move due to drying
shrinkage while also providing vertical load transfer. This means the dowels need
to be able to slip away from at least one side of the joint (bond breaker applied), and
ideally also have space to move parallel to the joint on one side of the joint. Walker and
Holland have published two excellent articles that sums up the use of plate dowels in
slabs (Walker and Holland, 1998, and Walker and Holland, 2007).
Joint transfer devices may be strategically located only at areas with the most
severe traffic conditions to optimize costs. A well-defined traffic pattern is generally
required.
The diamond dowels are steel plates that are laid flat and oriented at 45 degrees to the
joint. A sample installation for standard buildings is 1/4” thick x 4-1/2” x 4-1/2”, with a
given on-center spacing, and centered in the slab. There is a unique method of placing
these. A rigid plastic housing of adequate strength is first nailed onto initial concrete
edge form and gets included in the first concrete pour. The dowel is then slipped into
the housing for the second pour. This means that the contractor does not need to
provide holes in the formwork to accommodate a plate dowel placed in the first pour.
Another advantage of these dowels is the allowance for shrinkage movement in two
directions. The plastic housing is designed with an allowance for movement parallel to
the joint, and naturally the dowel slips out of the housing is there is movement
perpendicular to the joint. As mentioned, concrete panels shrink toward their center of
mass, so near the corners, the slab is moving diagonally. This leads to a component of
movement both parallel and perpendicular to the slab joint.
Fig. 37– Insert at edge of construction joint set into first concrete pour, with second
concrete pour to follow. Square dowels shown to be placed at a construction joint
placed in insert that was set into first concrete pour, with second concrete pour to follow
(https://www.pna-inc.com/products-designs/diamond-dowel-system)
ACI 360 recommends the following for dowel size and spacing.
Two types of dowels, round and tapered dowel baskets, are shown in Figs. 42 & 43.
The function of the dowel basket is to allow the control joint to open up under shrinkage
movements, but to maintain vertical load transfer across the joint. The tapered variety of
dowel has an additional benefit. First, note that the dowel is wider on one side, but that
the wider end is placed so it resides alternately on side of the joint, then the other. As
the concrete shrinks, the dowel holds to the side with the wider end, and when the joint
opens up, there is a gap that develops, allowing space to move for any movement
parallel to the joint. See Figs. 44 through 48 for dowel basket use and spacing, and
some construction details.
1) The use of a vibrator at joint devices is imperative for the devices to work
properly. If there are honeycomb voids, the devices may become loose, or there may
be weak sections in the concrete that are subject to higher stresses.
Isolation Joints
There will be many building components that will make the use of control joints difficult
to employ with complete regularity. Objects such as columns, bollards and doorways
will project through the plane of the slab, and with enough lateral movement of the slab
due to drying shrinkage or temperature effects, the restraint can induce cracking at the
margin of the object. The joints that provide relief at these objects are called isolation
joints, with an intentional gap and a compressible filler material, an example of which is
shown in Fig. 49. Re-entrant corners will exist where floor areas follow building
geometry and at transitions between adjacent areas that require discontinuities in the
slab. Special details may be required.
Joint Protection
Hard-wheeled traffic can be detrimental to slab joints. Over time, the action of repeated
cycles of forklift and pallet jack loading across a joint will tend to break the joint down.
ACI 360 notes two ways to protect a joint: armored angles or plates or joint filler.
Armored plates will be shown in a later course. Joint filler is discussed below.
Joint Filler
1) The most common type of material used is semi-rigid joint filler. Per ACI 360,
“Certain types of semi-rigid epoxy or polyurea are the only materials known to the
committee that can fill joints and provide sufficient shoulder to the edges of the
concrete and prevent joint breakdown.” Metzger/McGuire has an excellent
publication on the subject called Concrete Joint Filler. Per Metzger/McGuire, the
concept is to fill joints with an epoxy product that has “sufficient rigidity to avoid
deflection, protecting edges from impact damage”, “sufficient resiliency to absorb
impact”, and a low tensile and adhesive strength to allow the joint to move without
restraining it. Separation from one or both sides of the joint is expected. They also
provide guidelines for conditions for replacement such as joint spalling and when the
joint filler starts pulling out of the joint. Joint filler is used in both construction and
contraction joints. The joint filler is ony intended to be a partial depth installation.
Generally, the depth of joint filler is the depth of the sawcut or a maximum of 2
inches. Where contraction joints are large, silica sand can be used to choke off the
joint. At construction joints, a backer rod is used to define the joint filler depth.
2) Good joint filler materials have sufficient tensile elongation to perform in a ductile
manner.
3) The timing of joint filler ideally occurs as late as practically possible in the
construction phases of a building. It would even be best to wait until the ambient
operating temperature and humidity were achieved (e.g., during HVAC
commissioning). However, it is often the case that construction sequence dictates
jointfilling in the first 90 days. If this is the case, it is recommended that the joint filler
be inspected at one year or so later to assess its condition and be replaced as
necessary.
4) In cold storage faciltites, the joint filler should occur some time during the
temperature draw down sequence to allow so the temperature shrinkage has
minimal affect on the joint filler material.
Chapter 7 of PCA’s Concrete Floors on Ground provides guidance. Based on this, and
on the author’s experience, the concrete specifications for typical warehouses would
have FF/FL requirements in the range of:
FF/FL = 25/17 for normal lift traffic alone (no racks),
FF/FL = 35/25 for shorter racking of, say, 20’-0” final product height, and
FF/FL = 50/35 for taller or automated racking with, say, final product heights of 35’-0” or
greater.
The future course will also include information on special rack types such as automated
racking, as well as the design of cold storage facilities, and special detailing for unique
warehouse situations.
https://commons.wikimedia.org/wiki/File:VAN_(Very_narrow_aisle)_AGV,_courtesy_of_Egemin_Automation_Inc..jpg;
AGVExpertJS / CC BY-SA (https://creativecommons.org/licenses/by-sa/3.0)
Fig. 51 - Very narrow aisle racking
Finishes
The exposed surfaces of warehouse slabs take a lot of abuse from various activities.
Wood pallets with sharp wood edges and nails will abrade the concrete surface over
time. Hard troweling of the concrete paste at the surface will help with this. However, it
is good to have additional protection against surface wear. There are several options
available.
1) Sealer/Hardener: Silicate sealers
a. These are surface treatments that penetrate and harden the top layer of
concrete.
2) Integral Hardeners
a. Similarly, these harden the surface of concrete, but are an additive to the
concrete mix, so would provide protection for a slab that might experience
more depth of wear.
3) Shake-on or Shoveled-on Mineral Hardener (e.g., trap rock)
a. There are flat flakes of rock that are spread by spraying them on with a
shovel or by the use of a shake-on machine that drops them on the slab in
an even manner.
ACI 302 defines floor types by class in Table 2.1, and recommends finishes for classes.
Here is the excerpt for the two main warehouse (non-specialty) classes:
Most typical warehouses will fall in the Class 5 category, and will receive a hard steel-
troweled finish to densify and strengthen the surface, and then have a coating of silicate
sealer applied. ACI 302 and Chapter 8 of PCA’s Concrete Floors on Ground provide in-
depth guidance on this subject. This includes advising that normal practices of lower
water/cement ratios, good curing methods, and use of hard aggregates will aid the
abrasion resistance of concrete slabs.
Pre-Construction Meeting
One of the most important stages of a warehouse project is the pre-construction
meeting. It’s where all parties come together to discuss how the concrete slab
construction will be planned for and executed. Some topics to review:
1) Safety and equipment:
a. Screed types:
i. Laser
ii. Truss
b. Truck/pump access and pour methods
c. Pour environment
i. Air quality
ii. Seasonal issues
iii. Time of day/night
iv. Indoor/outdoor
v. Options to mitigate adverse conditions / thresholds for canceling
pours
2) Construction operation and logistics:
a. Other trades/materials
i. Soils/subgrade procedures and issues
ii. Underground plumbing or electrical ducts
iii. Insulation and heating coil contractors
iv. Reinforcing
b. Schedule – test pours, main pours, sequence
c. Working hours for pours
d. Control and construction joint locations
e. Roof in place or not
f. Vapor barrier locations
g. Joints at protruding elements (columns, bollards, etc.)
h. Joints to miss rack legs or base plate anchors & options at conflicts
i. Reinforcing to miss rack base plate anchors & options at conflicts
j. Mix designs
i. Pre-pour testing and timing
3) Quality control and assurance
a. Plan and parties for special inspections and testing
b. Any third party oversight
c. FF//FL and other special tolerances
d. Air, slump, temp, cylinders
References
ACI Committee 117, “Specifications for Tolerances for Concrete Construction and
Materials and Commentary (ACI 117-06),” American Concrete Institute (ACI),
Farmington Hills, MI, 2005, 70 pp.
ACI Committee 301, “Specifications for Structural Concrete (ACI 301-05),” American
Concrete Institute (ACI), Farmington Hills, MI, 2005, 49 pp.
ACI Committee 302, “Guide for Concrete Floor and Slab Construction (ACI 302.1R-04),”
American Concrete Institute (ACI), Farmington Hills, MI, 2004, 77 pp.
ACI Committee 318, “Building Code Requirements for Structural Concrete (ACI 318-
14),” American Concrete Institute (ACI), Farmington Hills, MI, 2014, 492 pp.
ACI Committee 544, “Report on Fiber Reinforced Concrete (ACI 544.1R-96),” American
Concrete Institute (ACI), Farmington Hills, MI, 1996, 66 pp.
Army Technical Manual No. 5-809-12 / Air Force Manual No. 88-3, Chapter 15 (TM 5-
809-1/AFM 88-3), Concrete Floor Slabs on Grade Subjected to Heavy Loads,
Headquarters Departments of the Army and the Air Force, Washington, DC, August 25,
1987.
ASTM Standard C309, "ASTM C309 – Standard Specification for Liquid Membrane-
Forming Compounds for Curing Concrete,” ASTM International, West Conshohocken,
PA
ASTM Standard D698, "ASTM D698 - Standard Test Methods for Laboratory
Compaction Characteristics of Soil Using Standard Effort (12 400 ft-lbf/ft3 (600 kN-
m/m3))," ASTM International, West Conshohocken, PA
ASTM Standard D1557, " ASTM D1557 - Standard Test Methods for Laboratory
Compaction Characteristics of Soil Using Modified Effort (56,000 ft-lbf/ft3 (2,700 kN-
m/m3))," ASTM International, West Conshohocken, PA
Azzi, Victor D., Ph.D, P.E., and Laird, Ralph H, “Load Carrying Capacity, Concrete
Slabs-On-Grade Subject to Concentrated Loads”,
https://www.structuremag.org/?p=5761, Structure Magazine, April 2008, Accessed
October 8, 2020.
Basham, Kim, Ph.D., PE. “Troubleshooting Concrete Cracks: Understand and Minimize
Cracking”, World of Concrete, February 4, 2013, Paradise, NV, KB Engineering LLC,
2013.
FEM 9.832: Basis of calculations for S/R machines, tolerances, deformations and
clearances in automatic small parts warehouses, Fédération Euroéenne de la
Manutention, 2001.
The Hard Hat Guy - Reviews from the Construction Site, www.thehardhatguy.com/best-
power-screeds/. Accessed August 2020.
Holland, Terry, D. Eng., and Groom, Jeff, PE. “Concrete Basics IV: Hardened Properties
and Durability”, World of Concrete, February 6, 2013, Paradise, NV, Terence C. Holland
and San Juan Construction, 2013
Holland, Jerry, PE, “Proper Use of Slab Reinforcement Systems: Avoiding Myths and
Misconceptions”, World of Concrete, February 7, 2013, Paradise, NV, Structural
Services, Inc., 2013
International Code Council. International Building Code (IBC 2018). Falls Church, VA,
2018.
Meyerhof, G.G., “Load Carrying Capacity of Concrete Pavements.” Journal of the Soil
Mechanics and Foundations Division, Proceedings of the American Society of Civil
Engineers. Vol. 88, June 1962 pp. 89-166.
RMI (Rack Manufacturers Institute) MH 16.1, (2012) “Specification for the Design,
Testing and Utilization of Industrial Steel Storage Racks.”
Shentu, L., Jiang, D., Hsu, T. (1997). “Load Carrying Capacity for Concrete Slabs on
Grade.” Journal of Structural Engineering, ASCE, January 1997, pp 99-103.
Tarr, Scott M., and Farny, James A.; Concrete floors on ground. Portland Cement
Association. Skokie, Ill., 2008.
Technical Report 34 (TR 34), “Concrete Industrial Ground Floors - A Guide to Design
and Construction“, The Concrete Society, 2003.
Vukobratović, Nino; Barišić, Ivana; Dimter, Sanja, “Analyses of the Influence of Material
Characteristics on Pavement Design”, Influence of Material Characteristics on
Pavement Design by Analytical and Empirical Methods, No. 14, 2017, pp 8-19.
https://doi.org/10.13167/2017.14.2.
Walker, Wayne W., and Holland, Jerry A., “Plate Dowels for Slabs on Ground”,
Concrete International, V.20, No.7, July 1998, pp. 32-35.
Walker, W.W. and Holland, J.A., 1999, “The First Commandment for Floor Slabs: Thou
Shalt Not Curl Nor Crack… (Hopefully),” Concrete International, V. 21, No. 1, Jan.,
pp.47-53
Walker, W.W. and Holland, J.A., 2001, “Design of Unreinforced Slabs-on-Ground Made
Easy,” Concrete International, V. 23, No. 5, May, pp.37-42
Walker, Wayne W., and Holland, Jerry A., “Performance-Based Dowel Design”,
Concrete Construction, 2007 World of Concrete Official Show Issue, January 2007, pp.
1-8.
Ytterberg, R.F., “Shrinkage and Curling of Slabs on Grade, Part I – Drying Shrinkage”,
Concrete International, V. 9 , No. 4, Apr. 1987, pp.22-31.