Home Blog Page 2

Over-strength Design Requirement (Ω = 2) in ASCE7 Specification for Baseplate and Anchor Bolt Seismic Design

The ASCE7 Specification is the document that the International Building Code refers to for applied loading for design of structures. Storage rack are covered in chapter 15 of the ASCE7. Section 15.5.3.2 of the ASCE7-10 Specification requires a modification to the ANSI/RMI Specification that includes the following statement:

Design forces that include seismic loads for anchorage of steel storage racks to concrete or masonry shall be determined using load combinations with over-strength provided in Section 12.4.3.2 of ASCE/SEI 7. The over-strength factor shall be taken as 2.0.

This basically means that the anchors are to be designed for double the computed uplift effect or ΩE where Ω = 2. This requirement would mean baseplates and anchors would have to be upsized to the point where the column base design is impractical. There is an option that is also contained within the ASCE7 Specification (Section 12.4.3.1). It reads as follows:
Emh = Ωo Qe
Qe is the effect of the horizontal forces and Ωo = the over-strength factor (Ωo = 2).

EXCEPTION: The value of Emh need not exceed the maximum force that can develop in the element as determined by a rational, plastic mechanism analysis or nonlinear response analysis utilizing realistic expected values of material strengths.

For the design of a baseplate and anchors the element being referred to is the baseplate. So a simpler step-by-step way to state all of the above would be:

Step 1: Compute the seismic uplift effect at the base, Qe. The baseplate and anchors must always resist at least this amount of uplift. The total uplift the anchors are required to take considering over-strength would be the lessor of:
a.) The uplift required to develop a plastic mechanism in the baseplate but not less than 1 x Qe.
b.) Double the uplift effect (2 x Qe.)

To comply with the over-strength design requirement the rack designer must choose either a.) or b.), above. If the computed seismic uplift effect (Qe) is small enough the designer will often choose option b.), then just make sure the anchors and baseplate are capable of carrying 2 x Qe. However, when the computed seismic uplift effect is larger the increase in baseplate and anchors may be extreme and option a.), will be chosen. By careful configuration of the baseplate and anchors a baseplate can be selected so a “plastic mechanism” will develop at an uplift value that is between 1 x Qe and 2 x Qe.

For example:
A storage rack is evaluated for cross-aisle seismic and the computed seismic uplift effect Qe = 5 kips. The designer chooses option a.). The designer configures the thickness of the baseplate geometry of the baseplate, the location of the anchors, etc. so the baseplate develops a “plastic mechanism” at an uplift of 6.0 kips. Based on the exception in ASCE7 12.4.3.1 anchors may be selected to resist 6.0 kips of seismic uplift rather than the 10.0 kips that would be required if option b.), had been chosen. The baseplate strength must be computed using the expected yield strength rather than the specified minimum yield strength.

Often in order to achieve a plastic mechanism using option a.), the thickness of the baseplate is reduced from the normally very thick baseplates that had been traditionally specified before this over-strength requirement was introduced into the ASCE7. The initial reaction to the thinner baseplate for those who have seen thick baseplates for years is to question the reason for the reduced thickness because using the thinner plate seems counter-intuitive.

The purpose of the over-strength requirement is to avoid brittle fracture of the anchors or the concrete slab region surrounding the anchors. Seismic energy is best absorbed by ductile elements rather than brittle elements.

The next edition of the ANSI/RMI Specification hopefully will address this over-strength issue so the ANSI/RMI exception can be removed from future editions of the ASCE7. For further questions regarding this requirement contact UNARCO Engineering.

 

Upright Table Interpretation

Customers always want to get the highest upright capacity with the lowest possible gauge so they can keep their costs down. Sometimes they get creative and interpret the upright capacity tables in ways they shouldn’t…

Here is a question that came up recently –

For a single, stand-alone rack bay is it acceptable to use half of the bay load and then select the frame from the frame tables since the frame only takes half the bay load rather than the full bay load that an interior frame would take?

No. The frame tables give the acceptable frame load for the frame when it is used as an interior frame in a continuous row (for low seismic areas). While it is true that an end frame in a row or a frame in a stand-alone bay only has half the applied load, these frames only have one beam framing into the top of the column span. The usual layout would have two beams framing into the top of the column span so the values in the table do not apply.

It is acceptable to use the full bay load and the tables for the end frame or the stand-alone frame selection. If there is any doubt as to how to select the frame, consult the engineering department.

How to Best Protect the Truck-Aisle Column

Fork truck traffic down the aisles can be the biggest threat to pallet rack uprights. The bottom portion of the truck-aisle column is the area of the pallet rack structure that is most likely to be damaged. An excellent option, particularly when the fork truck has outriggers is a welded bottom column straddle bumper. If racking is already installed, there are also many available bolt-on options for rack column protection. This option can be improved upon by making the wise decision to install a second anchor for the truck-aisle column.

If there is only the one anchor and the bumper (or column) is struck from the side or a diagonal direction, the entire bottom of the column will want to twist or spin about the single axis that is the lone anchor. The installation of a second anchor helps the column resist this tendency to twist. For this reason, a more complete and recommended solution to adding a bottom front bumper is to also add the second anchor to the front column.

In some applications, such as higher seismic regions, there may already be a requirement for more than one anchor in the aisle column. In these applications the two (or more) anchors used should be adequate to resist bottom column twist. However, in non-seismic areas, you should always consider adding a very cost-effective extra anchor even though local building codes do not require it. It is “cheap insurance.”

“Checker-Boarding”: Removal of First Beam Level in Every Other Bay

At UNARCO we often get asked if users can remove the first level load shelf in alternating bays to store taller pallets on the floor. Rack users think that removing a beam level in every other bay will allow them to have a shorter design column span and they can use the corresponding frame capacity from the manufacturer’s tables that goes along with that shorter column length.

This is a bad assumption because actually, the frame capacities listed in the manufacturer’s frame tables are based rotational stiffness provided by two beams connecting to the columns at the top of the column span not just one beam connecting at the top of the column span. Removal of every other beam significantly diminishes the story stiffness (ability to resist side-sway buckling of the entire system) and could lead to a story collapse of the entire rack row. Similarly, removing every other beam at the first story level of a high-rise building could have a devastating impact on the safety of the building.

The follow-up question that comes next naturally is, “How about removing the shelf in every third bay?” When the answer is again NO, they ask, “then, how come interior tunnel bays are permitted?” When the row is long and one level is removed in one bay close to the middle, the impact on the entire story stiffness of the rack row is minimal. If there are more tunnels the impact can become more significant.

So what are the options?
It is quite acceptable to “checker-board” the rack at the top by removing the top shelf in every other bay. Of course the logical option is to initially design the column for the longer column span, that is, from the floor to the second shelf level. If the rack system is already in place, it might be too late. Also, this option will require a significant increase to the column section and possibly the base plate and connections of the rack. A third option, if pallet loads are not going to relocate often is to segregate the tall loads and put them in a section of the rack together with a rack frame that is designed for these taller loads.

What is often thought as a simple solution, is actually not so simple at all and the myth of removing the lower level beams can lead to system failure. It is always a good idea when making system changes to check the entire design with engineering.

When is a 4” Pallet Rack Column a Better Choice?

UNARCO, as well as other rack companies, offer both 3” wide and 4” wide columns. Often, a 3” column is quoted out of habit without considering the potential benefits of using a 4” wide column. The racking industry is so used to a 3” x 3” design sometimes we fail to look at stronger, more economical options.

4”x 3” closed tube columns that are 13 ga. steel thickness (t = 0.083”) have a moment of inertia of about 2.52 in^4 where a 3”x3” closed tube column of the same thickness has a moment of inertia of about 1.27”. This means that for the extra 1” of column width you are essentially doubling the stiffness of the column. The table below shows a comparison for a

90” column span for these two UNARCO column sections.

Column  wt./ft.  Ix  Frame Cap. L = 90”
 4×3 – 13 ga.    3.75 #/ft.  2.519  26,019 #/Frame
 3×3 – 13 ga.    3.19 #/ft.  1.27  13,809 #/Frame
 (Unarco II punching)

For the case shown above, the capacity of the frame nearly doubles for a column weight difference of only 0.56#/ft. If the 3 x 3 frame selected needed to be 12 ga. its weight per foot would exceed the weight of the 4 x 3 13 ga. and its strength would still be less.
Here are some guidelines as to when to look at a 4” wide column:

1.) On pallet racks, pallet flow racks, Push Back racks or any beam-frame style rack when the column spans exceed about 72”, or anytime the 3 x 3 column gauge selection is a heavier gauge. When the 3 x 3 column is 10 ga. the 4” option is almost always more economical.

2.) On Drive-In racks greater than 22 feet to the top tie.

3.) On Pick Modules with more than one elevated floor. These tend to have longer column design lengths.

Many Pick Module racking designers and engineers have made the mistake of installing a module using 3” columns and have come back at a future time and made the request to re-configure or remove lower shelves only to find out that when they do so the column strength is not sufficient.

Another problem with building Pick Modules out of 3” columns is that “sway” beams often have to be installed to keep the column design lengths lower. These “sway” beams cannot be removed. Very often, the cost of the sway beams would have covered the upcharge for the 4” column. Many dealers prepare customer layouts using 3” columns and then later find that a 4” column would have been the right choice providing the customer with a stronger pallet rack design at a lower price.

In general, any application where the column span design length is longer than normal, a 4” column should be considered for both added story stiffness and economy.

Tying Racks to the Building Wall or Roof

Warehouses often move rack around from the originally designed layout to help store items more efficiently or help SKU adjacencies. Moving pallet racking is not always safe, especially when double runs of rack are changed to single runs of rack and components of the system such as back-to-back row spacers are removed.

In many cases, Warehouse Managers may decide to tie single-runs into a wall instead of the existing racking that it was designed to tie into. Is it a good idea to tie a pallet racking system to the building structure or walls? This question will often come up when the height-to-depth ratio of the single row of pallet rack along the wall exceeds the RMI limit of 6 to 1 or when a rack user wishes to use a set-back or slope-leg frame detail in a single row along a wall.

The latest RMI Specification discourages the use of wall ties because the building and the storage racks are each designed with their own structural system to resist seismic forces. When the two systems are tied together it can change the behavior of both systems. Forces from the racks are applied to the building through the wall ties and forces from the building are applied through the ties and into the storage racks. Technically, a structural analysis of the two systems as a combined system should be carried out if the two systems are tied together which can be a very expensive and complex engineering analysis, and sometimes cannot always provide an exact answer.

Wind forces into the building can also make their way to the rack through the ties. If the rack is tied to the roof, snow and rain loads, or roof wind suction can also cause force on the rack system. There is a lot to consider and analyze.

Unless your system was clearly designed to tie into building structures by a qualified structural engineer, it is recommended to avoid wall ties. If a single row height-to-depth ratio exceeds the 6 to 1, there are other design options available which include a larger baseplate and a stronger anchoring scheme or cross-aisle ties to tie the rack into another row that is across the aisle. Both of these options will help the system resist overturning and should be considered before the use of wall ties.

Damage to Upright Frame Bracing

Often times walking through the rack warehouse damage to frame bracing can be seen. Many believe that damage to bracing is not a problem if the columns are not damaged. This is not true. Damage to frame bracing can cause a rack frame to collapse even if the column seems to be undamaged. The column analysis relies on presence and structural contribution of the bracing.

In general a frame-beam rack structure has two very different structural systems that enable the rack to stand up and support the loads. In the down-aisle direction strength and stability are provided by the strength and stiffness of the column, the beams, the beam-to-column connections and the fixity of the base. This structural system is known as “semi-rigid” framing.

In the cross-aisle direction, strength and stability are provided by tension-compression strength of the frame bracing, the columns and also the anchorage. This structural system is known as “braced” framing. If a brace is damaged or lost for any reason the frame weldment can lose its integrity and ability to withstand cross-aisle force. This would be similar to a bridge truss with vertical or diagonal members removed. The bridge span strength would be compromised.

Part of any rack inspection should be to observe and report any damage to the bracing or their connections as well as column or beam damage. This damage should be repaired or the frames replaced. The safest warehouses utilize a regimented inspection program where all warehouse employees are trained to spot and report damage immediately. It is better to question whether a damaged brace has an impact on the racking system than to have damage go unreported and find out through a failure how important the structural integrity of the bracing is to the pallet rack as a whole.

Sloped Floor Pallet Rack Installations

Pallet Rack users often do not find out until the installation phase of a project that the warehouse floor is not level. Even with the uneven floor, the rack must still be installed plumb and all of the beams must be level enough so they seat properly.

Whether the slope of the floor is in the cross-aisle or down-aisle direction, shims are normally used to achieve a plumb install of the rack frames.

The height of the shim-packs should not exceed the vertical increment of adjustability of the frame. When this height is exceeded the installer can always just use the next increment of adjustment.

Example: A teardrop rack has 2” of vertical adjustment. If the shim requirement for level is 2-1/8” the installer would shim 1/8”, and use the next hole up on the column to level.

Pallet Rack Beam Infinitely Adjustable Angle
Infinitely adjustable beam connector

Some bolted styles of rack connectors can allow the rack to travel with the floor slope in the down-aisle directions to varying degrees. An “infinitely adjustable” angle on the beams can help with plumbing adjustments and make shelves level. As previously stated, all of the beams must be properly connected and all frames must be plumb.

What should be done about high shim packs?
The RMI Specification defines three possible conditions:

1.) For shim packs that are less than 2 times the diameter of the anchor, there is no requirement to “lock” the shims together by welding or installing a second anchor.
2.) For shim packs that are between 2 times but less than 6 times the diameter of the anchor, there is a requirement to “lock” the shims together by welding or adding a second anchor.
3.) Shim packs that are greater than 6 times the anchor diameter are not permitted.

One very important thing to remember is that longer anchor bolts may be needed to achieve the required anchor bolt embedment for those cases where shim stacks are used.

A common problem with shim packs that are too high and not welded or fastened with a second anchor is shim “spin-out.” This can lead to instability of the racking system.

Your Questions Answered on Beam Loading and Beam Deflection

Does it matter how I place pallets on the rack?
Manufacturer’s beam tables are generally based on uniformly distributed loads. Both the capacity and the beam deflection will vary if the load is not uniformly distributed on the load shelf beams. A short list of things that can cause non-uniform loading are:
1.) Feet or runners on the load that cause point loading on the beam.
2.) Loads that have non-uniform weight or mass over their projected area.
3.) Loads that are not symmetrically placed on the shelf either in depth or width.
4.) Loads that are undersized for the shelf. (These loads can be placed towards the center of the beam rather than towards the rack frame. This will aggravate the stress and deflection of the beam.)

How are beams sized by the rack designer?
Beams are checked for strength (stress) and deflection (serviceability) and the limit that is reached first will govern the rating of the beam. Beams that are deep relative to their span will reach their stress limit first and are referred to as “stress governed”. Beams with a longer span that are shallow relative to their span will reach their deflection limit first and are referred to as “deflection governed”. The RMI deflection limit for a normal pair of pallet rack beams is L/180 or the span divided by 180. A 96” beam will have a deflection limit of 96/180 = 0.533”.

Why is the RMI Deflection limit L/180?
This deflection limit is an aesthetic limit that was set forth many years ago. It is based on what was judged to be the normal visual comfort level for deflection. Excessive deflection will and should cause unease for those in and around the rack. It is worth noting that some believe they should see no deflection and are alarmed when they witness any deflection. To require beams that exhibit only very slight deflection would be very expensive and require beams that are very deep relative to their spans.

Are there some cases where L/180 deflection is too much?
Yes. Often times where automatic equipment is involved a much tighter deflection limit is specified for the load beams. Also, catwalk beams that support walking aisles should be designed for less deflection (L/240 is often used).

How do I properly measure the deflection of a pallet rack shelf beam?
The easiest way to get a fairly accurate deflection is to stretch a string across the bay that is some distance below the beam (usually just below the connectors). The string is taught and is adjusted up and down at the ends so that the distance from the bottom of the beam and the string is the same at both ends. Mark the centerline of the beam. Symmetrically place the loads (weighed loads) on the beam in both the front and back direction and the side-to-side direction. The loads should be uniform in weight so the front and the rear beams are equally loaded. Measure the distance from the bottom of the beam to the string at mid-bay and subtract this from the distance from the bottom of the beam to the string at the ends. This is your beam deflection. Bear in mind that there is measurement error here as in any other procedure.
The reason to use the string in this manner is that it removes error that may result if you just measure from the floor with a tape measure. Doing this may cause error if the floor is not level. Another reason for using this method is that any downward displacement of the connection is also eliminated from the deflection. The deflection test is to determine only the actual bow in the beam cause by the loading. There are also more high-tech ways of checking deflection with lasers, etc.

What should my deflection be if my applied load is less than the rating of the beams?
The deflection should be relative to the deflection capacity of the beam if the load is less. For example, if a 96” shelf with a stress capacity is 5000# per pair and a deflection capacity of 5500# per pair, is uniformly loaded with 3000# per shelf, the anticipated deflection at 3000# would be: (3000/5500) x 0.533” = 0.291”.

Would added yield strength reduce deflection?
No. Deflection is not a function of yield strength. Added yield strength may increase the stress capacity of the beam but not the deflection capacity.

Would added steel thickness reduce deflection?
Yes. Added steel thickness will increase the stiffness of the beam and reduce deflection. A more economical way to reduce deflection is added section depth because a beam section’s ability to resist deflection is an exponential function of its depth and only a linear function of its thickness.

Seismic Separation and the Pallet Rack Layout

In the last five years, seismic separation in racking layouts has become increasingly enforced and a section on seismic separation has been written into the 2012 RMI specification. Seismic separation is the required horizontal distance between the building columns or walls and the racking components. To compute the seismic separation requirement for a rack project in an elevated seismic risk area, the seismic displacements of the rack and the building need to be known or calculated. Once these values are determined, the required seismic separation can be computed. These separation distances will be different for the down-aisle and the cross-aisle direction of the rack and may have more impact on warehouse design in areas with higher ground accelerations.

The required seismic separation will usually be greater than a typical rack layout would allow for a distance from the building walls or columns to the rack. Most engineers allow as much aisle space possible to help forklifts maneuver the warehouse. In higher seismic areas, the building officials may require that the layout be changed so the required separation distances can be established. This can have a significant effect on the rack layout and may cause problems when the rack user has to increase flue space and reduce forklift aisles. Back spacers may also fall too close to frame lines in the down-aisle direction and have to be re-sized or adjusted. A typical proposed layout may only have 2-3 inches between the building column and the rear pallet rack beam but when the seismic separation is computed, it can be up to 12 inches or more. The warehouse layout can be drastically different from the original design once these values are considered and pallet placement locations are usually impacted.

What is the purpose of these requirements?
The purpose is to avoid the collision between the rack structure and the building structure in the event of an earthquake. Because the rack and the building will both move independently during the earthquake, both must be considered in the calculation.

What should be done initially?
The rack user should be warned about this potential issue and that the requirement may cause the layout to be altered once the final design is complete and the separation requirements are determined. The building owner will need to try to find out what the seismic drift of the building is in each direction. If this value is not available, there are default values in the RMI Specification that can be applied. The rack engineer will calculate the amplified seismic drift of the rack in the cross-aisle and the down-aisle directions. The racking layout will then be adjusted to include the new values of separation.

In the end, bay configurations will need to accommodate the movement of the pallet rack and the building during an earthquake, so working seismic separation into the equation as early as possible may stop the headaches of a complete layout change before the project is too far along.