Home Blog

Load Capacity Requirements in the new RMI Storage Rack Standard ANSI MH16.1-2021

The revised American National Standard for storage rack, ANSI MH16-1-2021, was recently published by the Rack Manufacturers Institute (RMI), and includes new requirements for calculating the load capacity for a pallet rack application. The previous standard, published in 2012, calculated capacities using only the column unbraced length.  New guidance in the 2021 revision uses several additional factors that affect the stiffness of the pallet rack structure to calculate the capacity.

The capacity of the frame will depend on nine factors beyond just the length of the unbraced column span, including:

  • Average load to maximum load ratio
  • Beam-to-column connector stiffness and strength
  • Beam stiffness
  • Column stiffness
  • Base plate and anchorage detail
  • Site seismicity
  • Number of storage levels
  • Aspect ratio of the frame, or height-to-depth ratio
  • Warehouse or Retail environment

In order to determine if the global stiffness of the rack assembly meets the new requirements, the designer will need to perform several calculations using the above nine criteria. Rack users need to understand that a frame capacity calculated in accordance with the 2021 standard may be less than the capacity calculated using the 2012 standard, all things being equal. 

Users should be careful using traditional upright capacity tables because there are so many other factors to consider now.  Tables can be used as an initial guide, but more calculations must be completed before a frame capacity is accurate and final.  With a reasonable average load assumption and stiffer connectors at lower beam levels, in most cases the rack design will be similar to capacities calculated under 2012 standard.  In some cases, no rack design changes will be needed. 

One important, key factor to determining a capacity with a greater chance of correlation to the 2012 standards is the AVERAGE LOAD DATA.  If the average load data at the facility is available, and connectors with adequate stiffness are selected, then the calculated frame capacity may be close to the capacity calculated under the 2012 requirements.  In such cases, the average load data provided will become part of the design assumption and must appear on the load application and rack configuration (LARC) drawings and the load plaques.

*** Please note that the RMI does not allow a load that is less than 67% of the maximum load for the calculation. ***

While the new methods make frame selection a bit more complex, there are also advantages. One such advantage can come from identification of an average product load per bay, as opposed to a traditional maximum product loading calculation.  Many storage racks utilize an average loading that is less than the maximum loading in all storage locations. This may be due to empty positions with no load, or the number of bays with lighter product pallets. Failure to identify an average product load will likely require the use of stiffer connections or an increase in the column strength to comply with the 2021 revision requirements compared to the 2012 requirements. 

More information will definitely be needed to determine a calculated capacity and this will cause users to get more involved in the design.  Once consumers start looking at the nine factors listed above, they will be more educated for future projects and start looking for average load data, so they are not forced to use the higher criteria, which will usually come with a higher price tag.

The consideration of all of the above factors makes working with the rack designer even more important than ever.  With a thorough understanding of your rack application, a rack designer will be able to select the most economical frame that still complies with the 2021 standard.  UNARCO Engineers can help navigate the new code and help with an economical design to increase efficiency in any warehouse or distribution center.

Anchor Bolts: Nominal Embedment vs. Effective Embedment

There is often confusion regarding what is meant by nominal embedment of an anchor and what is meant by effective embedment of the anchor. The illustration below will help with this.

Note that the nominal embedment is the distance from the floor surface to the bottom tip of the anchor prior to tightening (see “before” picture above for hnom).  When an anchor is designated:  ½” dia. x 3 3/4POWSD1 on the plans the 3 3/4 is the nominal embedment.  After the anchor is tightened the remaining embedment may be somewhat less due to the upward travel of the bolt into the wedging sleeve.  Anchor bolt inspectors should understand the difference and should not try to enforce nominal embedment after tightening.  If the “after tightening” embedment is less than hef then the embedment is not to specification.

The “effective” embedment hef (shown in the “After” picture) is the value used for the concrete breakout calculations.

Anchor Bolt Embedment Tables

On projects in elevated seismic zone where there is significant calculated uplift force on the anchors, deeper embedment depths may be required for the anchor bolts. For these projects it is important that the thickness of the slab is adequate to properly achieve the necessary anchor bolt embedment. For a given anchor bolt diameter and embedment there is a chart in the anchor bolt product manual and in the anchor bolt ESR report that gives the required slab thickness needed.

Here is an example:
Project requires ½” diameter Powers SD2 anchors with 4 ½” minimum embedment.
Since 4 ½” embedment is not in the table we will extrapolate:
For 3 ¼” embedment the required hole depth is 4” and the required slab thickness is 5 ¾”.
For 4 ½” embedment, add 1 ¼” (4 ½” vs 3 ¼”) to each value:
Minimum hole depth = 5 ¼”. The required slab thickness = 7”.

If the slab is only 6”, that will not work for the 4-1/2″ anchor. The slab either needs to be a minimum of 7”, or the base and anchorage needs to be changed to a lesser embedment. In summary, the required anchor bolt embedment depth must be coordinated with the slab thickness so proper installation of the anchor bolts can be achieved.

Concrete Slab Importance in Rack Anchor Embedment

It is important to know the slab thickness of the rack floor because there is a minimum slab thickness that is required based on the required nominal embedment of the anchor bolt.

For example, if your rack design required the anchor bolt nominal embedment to be 4-5/8” the slab must be at least 7” thick (See minimum member thickness in table at arrow). The minimum slab thickness is required so the installer will not tend to drill through the slab when installing the anchors. For cases where the slab is not thick enough, the anchor detail must be changed to require less nominal embedment. This is mainly a concern when there is seismic uplift demand on the anchors. Below is a cut from a table where the minimum slab thickness for given embedment is shown:

For normal anchors (L = 3-3/4” or 4”) that require only 2-1/2” nominal embedment, the floor will generally be plenty thick because most industrial floors are 6” thick or more (Caution – There are some that are 5” thick).

It will be helpful (and sometimes very necessary) to also know the strength of the concrete of the slab-on-grade. This can often be found on the building drawings. The building drawings may also show the rebar that may be present in the slab. This information is important because the rack designer and installer may know to what extent the rebar interference may be a problem. When there is a lot of rebar, it may be helpful to provide a larger baseplate with alternate anchor holes so the installer can re-drill if he encounters rebar.

Below is an example of a baseplate with alternate anchor holes:

One additional item from the building drawings that is also important is the location of the joints in the slab. The rack should be laid out so the anchor bolts are not too close to these floor joints (preferably at least 1.5 times the embedment depth of the anchor). This does not apply as strictly to shallow saw cuts in the slab although these should also be avoided when possible.

Adding Frame Extensions

Anytime a rack user wants to change their bay profile or loading, the rack should be reviewed by a qualified engineer to ensure that the new proposed configuration is still RMI compliant. This includes adding frame extensions that increase the height of the storage rack.

Adding frame extensions can be a problem even if the total load in the bay is not increased. When shelves are moved upward (raised in elevation) the strength of the framing structure is affected. Also the addition of the frame extension can cause the height-to-depth ratio (aspect ratio H/D) of the rack to increase if shelves are added above the splice point. This addition may cause the H/D ratio to exceed 6-1 ratio. When this ratio is exceeded extra measures for stability are required by the RMI Specification.

Adding frame extensions so additional shelves can be added that causes the total bay load to increase can obviously be a problem if the strength of the existing frame is not great enough to handle the added load. In elevated seismic risk locations, raising the center of gravity of the loading on the frames by adding load and bay height may cause the anchorage to be overloaded during a seismic event.

In summary, anytime the bay configuration is changed or load is increased the rack should be re-evaluated by a qualified engineer.

Exception: If frame extensions are added only for product fall protection or to support product fall protection netting and the load beam elevations are not changed a design review may not be needed.

Shim Installation: Cross-Aisle Plumb

The RMI Specification requires that loaded pallet racks be out-of-plumb no more than H/240 where H = the distance from the floor to the top shelf level. To put this in simpler terms this allows for ½” out-of-plumb for every 10’ of height. This limit is for a loaded rack. The actual installation tolerance for plumb should be much tighter because any out-of-plumb on an empty rack will increase once the racks are loaded. The installer should make the racks as plumb as reasonably possible so the final out-of-plumb stays within the RMI limit for the loaded racks. Most installers will want the installed empty rack out-of-plumb to be less than half of the H/240 limit so they are not called back to the site once the rack is loaded.

Before attempting to plumb the racks down-aisle the racks should be installed plumb in the cross-aisle direction. This is normally done using shims. Frames with greater aspect ratios (Height-to-depth ratio H/D) will need to have thinner shims to control the top movement of the frame when a shim is added. For example if the aspect ratio of a frame is 6 to 1, a 1/16” shim will move the top of the frame 3/8” (6 x 0.0625”). If a 1/8” shim is used the top of the frame will move twice this much or ¾” (6 x 0.125”). Having only the thicker shim will make it more difficult to plumb the rack. The manufacturer should provide shims that take into account the top movement as it relates to the detail of their row spacer.

When the rack frames are set back-to-back and are to be connected with row spacers the frames are normally set the correct distance from each other at the floor and are normally installed following a chalk line on the floor. In a perfect world that flue space would remain exact for the full height of the two frames. In reality though the flue space may either diminish or increase at higher elevations. This is why the cross-aisle movement of the top of the frames caused by the addition of shims becomes extremely important. A diminishing flue dimension may be caused by a dip in the floor in the flue area.

Some manufacturers make row spacers with slotted holes to allow for minor variation to the flue distance. Other manufacturers will have a row spacer that must fit in between the faces of the two frames at the top. If the flue dimension diminishes it may not be possible to connect the back connector because it will not fit between the faces of the columns at the top. A common installation mistake is to loosen the anchors and pound in the back connector. Doing this can actually cause the column at the flue to no longer be in firm contact with the floor. (See Figure A below). When this happens things may be fine until the racks are loaded. Once the racks are loaded the columns in the flue are forced downward and the frames may compress and buckle the top row spacers in the down-aisle direction which can force the racks out-of-plumb down aisle (see figure B below). To prevent this it is actually better if the flue dimension at the top is either the same as it is at the bottom or slightly greater, rather than slightly less. The back connector at the top should easily fit (with all columns in firm contact with the floor) and not have to be forced in.

Uneven uprights

A good procedure to prevent this from occurring would be:

1) Set up two rows without row spacers.
2) Properly shim to cross-aisle plumb so flue dimension at top is greater than or equal to flue dimension at the bottom and top back connector installs easily with all posts of rack frames in firm contact with the floor (Tighten anchors to ensure)..
3) Down-aisle plumb the rack after the cross-aisle plumb is complete.
4) Install back connectors.

Rather than purchasing (1) shim per frame that may be 1/8” thick it is wise to purchase (1) shim 1/16” thick and (1) shim 1/32” thick to give the installer more ability to “fine-tune” cross-aisle plumb. If the floor is suspect thicker shims can be purchased as needed.

Cantilever Racks New Design Requirements

Recently the first ever RMI Cantilever Specification was completed and adopted by the RMI (Rack Manufacturers Institute). Following this adoption the RMI submitted this document to ANSI for approval to become an ANSI (American National Standards Institute Specification) standard. Before ANSI approves a specification, the specification has to go out and be approved by a canvas committee that is made up of various interest groups that include manufacturers, consultants, users and those who may have a general interest in the document. At the conclusion of the review, all comments that are submitted are addressed prior to final ANSI approval.

During the final review of the Cantilever Specification, an engineer from the seismic engineering community realized that there were no special seismic detailing requirements for cold-formed steel cantilever racks. Due to the absence of these detailing requirements, he commented that greatly reduced R values needed to be required until these detailing requirements could be developed and met. Detailing requirements for structural steel can be found in the AISC 341 Specification. These requirements place strict limits on the b/t (width-to-thickness) ratios for stiffened and unstiffened web flanges to prevent local buckling and add ductility. These limits are well exceeded by most normally designed cold-formed cantilever racks. Below is a table from the new RMI Cantilever Specification showing the requirement for reduced R values:

RMI Seismic Design Parameters

Notice in the table above the cross-aisle R value changes from 3.0 to 1.0 for category D which essentially triples the amount of seismic force the rack is required to resist. The down-aisle R value changes from 3.0 to 1.5 which doubles the forces on the bracing. In column and base design for category D, the column and base design will be governed by the seismic load case. This means the cross-aisle moment will be tripled by this requirement. This will add significant cost to these structures and possibly make them impossible to design in some cases. The computed seismic uplift will also be extreme.

A reasonable question to ask would be, “Why didn’t the RMI just refuse to change the R requirement in their document?” The answer is two-fold. First, ANSI approval would not have been possible with the comment from the seismic community being ignored. Secondly, if the RMI was to be adopted by the building codes (ASCE7 and IBC) without the reduced R values, an exception probably would have been added to the ASCE7 stating, “Use the 2016 RMI Cantilever Specification except use the requirement for the lower R values for Category D and above.”

Another question would be, “What is the effect on Structural Cantilever racks?” Structural cantilever racks will have to meet the detailing requirements set forth in AISC 341. This will disqualify sections that exceed the b/t limits in AISC 341 TableD1.1. They also are required to have the column to base connection designed to include the over-strength multiplier (Ω) or the full Mn (Ultimate moment strength) of the column.
The first footnote in the table above does allow cantilever racks designed using cold-formed steel if they are designed in accordance with AISC 341. The difficulty here is that the b/t requirements require the plate elements to be so thick that the sections cannot be easily fabricated using tooling that exists in most fabrication shops.

Question: How do I know the seismic design category for my site?
To determine the seismic design category, the engineer must first determine the Sds and the Sd1 values for the site. To do this, the soil site class is needed (if available). If the soil site class is not available the ASCE7 requires that the default site class (Class D) be used. Using the soil site class and the Ss and S1 values obtained from the USGS website for the address of the site, the Sds and Sd1 values are determined in accordance with chapter 11 of the ASCE7 Specification. The storage rack engineer can quickly determine the seismic design category if the soil site class and the address are known. If the soil site class is not known the engineer will have to use D. It can be very advantageous if the soil site class is A, B or C. The table below (from ASCE7) then defines the seismic design category. For storage racks the Risk Category for the table would be the first column labeled I or II or III.

Question: What are my options if I find that my site is in seismic category D?
Since the seismic forces will be extreme, the design of the cantilever structure will likely need to change. More columns can be used to reduce the force per column. The structure may have to be shorter with fewer arm levels. Heavier loads could be required at low elevations in the rack and lighter loads at higher elevations.

Another option would be to look at the load lengths and use pallet rack for any load that can be short enough to possibly be stored on pallet type rack.
Compliance to the new R values given in the ANSI/RMI Cantilever Specification will significantly increase the cost of racks going into seismic design category D.

Note that another option is to choose stockier members (less slender elements) so that the width-to-thickness limits in Table D1.1 of the AISC 341 specification are not exceeded. This will allow more reasonable R values.

New Code Editions Increase Seismic Forces

More and more states are adopting newer codes- in particular IBC 2018, which references ASCE7-16.  An example of this is the 2019 Oregon Structural Specialty Code which went into effect on October 1, 2019 or the 2019 California Building Code which goes into effect on January 1, 2020.  The ASCE7 lays out the standards for seismic designs and in the new ASCE7-16 edition, there is a note that is not found in previous editions:

“Where Site Class D is selected as the default Site Class, Fa (site coefficient) shall be taken as 1.2”

This note presents a problem for seismic designs.  Previous versions of ASCE7 have allowed for the use of the table values when using the default site class.  The table values allow for a reduction in the site coefficient when you are within a certain range of Ss values.  Using previous codes as a reference point, this new note requires an increase in seismic forces as follows:

  • For Ss between 75 – 100, the increase is 0-10%
  • For Ss between 100 – 125, the increase is 10-20%
  • For Ss above 125, the increase is 20%

Due to the provisions in the ASCE7 that are mirrored in the RMI code, Site Class D is currently used as the default.  Table values that allow for a reduction to the seismic forces are permitted to be used only if you confirm Site Class D using geotechnical data. 

One solution is to ask the customer, building engineer, or building department for data specific to the job site.  The building drawing will list the site class.  Obtaining this information will allow for a more economic design.  Furthermore, if site class data is found to be A or B, the seismic forces will actually be reduced from the table values for Site Class D.  For example, if it is found that the jobsite is Site Class B the seismic forces can be reduced by 10%.

Another important piece of information is the slab depth and the concrete compressive strength.  Usually a customer, building engineer, or building department will have this information handy and can provide it.  This information helps greatly with the design of the baseplate and anchorage detail.

Drive-In Columns – Slenderness

The Drive-in rack configuration is unique because the fork-lift actually enters into the storage bay. For this reason, there cannot be a cross-bay beam except at the very top of the bay at the rear-most storage position which is beyond the travel depth of the truck. In the case of a double entry drive-in, where the truck enters from both sides and travels toward the middle, these cross-bay (anchor) beams may be in the middle area of the bay. For the columns without cross-bay beams, the load rails are supported by load arms that attach to the tall slender column.

Drive-in can be a very sound and reliable rack structure when the column is adequately stiff, but can be problematic or even dangerous when the column stiffness is not sufficient. The obvious problem that can occur when the column stiffness is inadequate is side-sway collapse. The less obvious problem is that for very tall slender column, the slightest lateral bump at mid-height can cause the column to deflect outward and the distance between the guide rails to increase, allowing pallets to drop from the rails.  this occurs because there is typically only about 1” of bite figured for a pallet that is side-shifted to one side between the guide rails. This problem can occur when a designer is tempted to use a slender column that “checks” because the customer’s pallet weights are not heavy, but the designer has not considered the problems mentioned above. This may not be as much of an issue when the load weights are heavier because the slender column may not even “check” for the heavier load scenarios.

The specifications have traditionally discouraged the use of compression members whose slenderness ratios (KL/R values) exceed 200 for these types of reasons. For a drive-in rack with a 3” column this ratio is exceeded for a symmetrically loaded drive-in column height in the 22’ to 23’ range. At this height, the design should go to a 4” column. Similarly, at about 31’to 32’, the column design should go to 5”. These are not absolute numbers, but vary depending on the center of gravity elevations of the applied loads and a number of other factors.

It is also generally more economical to choose a column that is not so slender.  For example, a 4” column may be a better and cheaper choice than trying to double a 3” column for a drive-in application because the stiffness (moment of inertia of the section) for a 4” column may be higher than the doubled 3” column and weigh less. For example, a C3 x 3.5 structural channel column doubled is 7# per foot and has a moment of inertia of 3.14 in^4.  The single C4x4.5 column weighs only 4.5# per foot and has a moment of inertia of 3.53 in^4.  The moment of inertia property is a measure of the section’s ability to resist deflection (or a measure of stiffness). From both a cost and a stiffness standpoint, the 4” column becomes the better choice.

Finally, aesthetics should also be a concern because a slender looking column can make those charged with working in and around the racks uncomfortable or unsafe.