The known bugs page was formerly hosted here and has since been moved to the MASS software documentation site:

Known Bugs in MASS

For questions about specific bugs, or to report a bug, contact mass@canadamasonrycentre.com

If you have ever designed a multi-storey shear wall and wondered why the moment resistance is less than expected, the reason is likely CSA S304-14: 10.2.8:

MASS automatically identifies shear walls that have an aspect ratio less than 1 and designates them as squat shear walls. Keeping all calculations and design results in accordance with the CSA Standards, it also correctly reduces the moment arm of all steel in tension when applicable which is why there is a reduction in moment resistance. While it is often the first reaction of many users to assume that this behaviour comes from a bug in the software, MASS is behaving as intended.

Multi-Storey Applications

What if you are designing just one element within a larger shear wall where the element has an aspect ratio less than one but the full shear wall does not? Is it correct to be applying the reductions from clause 10.2.8 to elements such as these? Consider the example below:

This example which was used in the Multi-Storey Shear Wall Design article demonstrates an instance where this clause comes into play. The entire shear wall itself is clearly not squat as it’s aspect ratio is 3.2. As it is loaded, it is behaving as a non-squat shear wall so it is not correct to be applying clause 10.2.8 to the design of an individual storey. In order to design this wall in MASS, only the individual elements can be modeled and designed separately. As you can see, the wall input into MASS on its own is designated as a squat shear wall which is where the Total Height input comes in handy: it allows the user to tell MASS that while an element may be “squat”, it should not be treated as such.

“Height” vs. “Total Height”

The scenario described above is the reason multiple height inputs are available in MASS.

Height refers to the vertical dimension of only the shear wall element being modeled while Total Height refers to the vertical dimension of the full shear wall assemblage, beyond just what is being modeled.

If Storey 2 is modeled in MASS without any consideration of the larger shear wall it is apart of, it is designated as being squat as it’s 4/5 aspect ratio is less than one. When the total height is changed to the full 16m, the aspect ratio used to apply squat reductions from clause 10.2.8 increases to 3.2 and the result is an improved moment resistance.

Impact on Design

How much of a change does this make to a shear wall design? Using the example from earlier, when designing using a 20cm, 15MPa concrete masonry unit, taking the total height into account means the difference between using No. 15 and No. 20 bars placed exactly the same. If using No. 15 bars for both designs, the squat version of the MASS file would need to go all the way from a 15 to 30MPa strength unit to compensate. Furthermore, if the masonry and reinforcement properties were both fixed to the same design, the difference in capacity can be seen on the interaction diagrams below:

Comparison of moment envelope curves for shear wall design both including and neglecting the total height

For the critical load combination (#15: 0.9D + 1.4W), this means that the moment resistance of the wall is reduced from 1333.5kN*m to 1111.5kN*m, or by 222kN*m, simply by not taking the aspect ratio of the full wall into account!

This effect is further demonstrated in the example below where 70% of the vertical reinforcement is concentrated on either end of the wall. This significant reduction in moment is a direct result of a reduced moment arm for the steel that is in tension and furthest away from the compression zone. Note that this design uses the exact same materials simply arranged differently.

Comparison of moment envelope curves for shear wall design both including and neglecting the total height

There is now a 330 – 430kN*m reduction in the moment resistance compared to the 200 – 275kN*m reduction observed when the reinforcement is evenly distributed. One thing to note for all of these comparisons is that the difference in moment resistance diminishes when the applied axial load approaches P_f,max.

For those curious, a comparison of the uniformly distributed reinforcement and concentrated end steel designs can be found by expanding the section below:

If you have any questions, please do not hesitate to call or email the Canada Masonry Design Centre.

The MASS software is a product of a joint partnership between CMDC and CCMPA. CMDC is the authorized provider for MASS Technical Support.

When software bugs are found, notifications are posted in the MASS website Known Bugs page, on the MASS welcome screen, as well as here on the CMDC website software blog in an effort to be transparent and keep all MASS users informed. This issue was found within our own office as a result of an inquiry from a MASS user that resulted in this post. The bug was found some time after and as a result, the fix was incorporated into MASS Version 3.0, as well as a modified copy of Version 2.2 specifically to address this issue (click here to read more about Version 2.2.1).

This post outlines the conditions required to trigger this error, the designs where this could have affected design results, the details of the bug itself, and how to check to see if this bug is present in any MASS project.

Jump straight to:

• Bug Summary
• Background information
• What types of designs might be affected
• How to check if a design is affected
• Example
• Our Response

Bug Summary

Under a very specific combination of conditions, MASS may calculate the critical axial load for a reinforced wall, P_cr using both (EI)_eff of 0.4E_mI₀ and Φ_er of 0.75, resulting in a P_cr value that incorrectly combines aspects of reinforced and unreinforced analysis.

This bug only affects designs where all of the following conditions are met:

1. vertical reinforcement is not in tension
2. load combinations with steel in compression also govern the design
3. the wall experiences slenderness effects but is not a slender wall (kh/t > 30)

If MASS designs a wall where the vertical steel is in tension, all of the software’s results are correct. The same is true if any of the other conditions listed above is not true. The unlikely combination of circumstances is the reason why most wall designs are not affected and also why this bug has not been discovered until recently. This has been addressed in MASS Version 3.0 as well as in an updated Version 2.2.1.

Background information

If a wall is reinforced, why is it a problem to use the reinforced reduction factor when it comes to slenderness effects?

There is no problem as long as everything else in the calculations is also treating the wall as if it were reinforced. The issue comes from the way in which MASS sometimes treats the wall as if it were unreinforced when the steel is not in tension. This is done intentionally as ignoring the steel allows the software to use CSA S304 chapter 7 clauses governing the analysis of unreinforced walls when it is beneficial. This also assumes that the addition of steel to an unreinforced (but still grouted) wall will not reduce its capacity. For a full explanation on RM/URM analysis for reinforced walls, click here.

Differences between treating the wall as reinforced vs. unreinforced

Reinforced analysis compared to unreinforced differs in two ways when it comes to determining slenderness effects of a wall:

1. Φ_e vs. Φ_er – Resistance factors for member stiffness used for slenderness effects
2. (EI)_eff – Effective stiffness for consideration of slenderness

Resistance Factors – Φ_e vs. Φ_er

There are two different resistance factors that are used in calculating the critical axial compressive load used for determining slenderness effects; Φ_e for unreinforced walls, and Φ_er for reinforced walls. Click the heading below to expand the CSA references from which these resistance factors are based upon.

As stated at the beginning of this post, the bug is that for reinforced walls, the reinforced reduction factor is sometimes incorrectly applied, using Φ_er rather than Φ_e. On its own, this would not be an issue, however, when combined with the reinforcement assumptions used in determining effective stiffness, there can be a “mixing” of assumptions that results in this bug.

Effective stiffness – (EI)_eff

Similar to the resistance factors, the formula used to determine effective stiffness for slenderness effects is a function of whether or not the wall is reinforced. Effective stiffness, or (EI)_eff, for unreinforced walls is 0.4E_mI₀ while (EI)_eff for reinforced walls is 0.25E_mI₀ (where the applied eccentricity is relatively low, below the Kern eccentricity). Click the heading below to expand the CSA references related to effective stiffness.

Reinforced walls with low eccentricities have an effective stiffness capped at 0.25E_mI₀. Compared to 0.4E_mI₀ which is used for unreinforced walls, there is a 37.5% reduction in stiffness simply for using the reinforced equation for the same wall design. Taking advantage of the unreinforced effective stiffness is what also mandates the use of Φ_e rather than the higher Φ_er.

What types of wall designs are affected?

As mentioned in the summary shown here,

• vertical reinforcement is not in tension
  • c is greater than d for any bar
• load combinations where c exceeds d also govern the design
  • Load combinations with higher axial loads are more likely to have the neutral axis exceed d while load combinations with both low axial load applied with high bending moment are typically closer to a wall’s interaction diagram envelope curve
• the wall experiences slenderness effects but is not a slender wall
  • Slenderness ratio, kh/t, is greater than 10 – 3.5[e₁/e₂] specified in S304-14: 10.7.3.3.1 but less than 30 where axial load is governed by S304-14: 10.7.4.6.4

The only wall designs affected by this bug are those which have axial loads so high that the steel is no longer in tension AND experience slenderness effects without being classified as slender walls (S304-14: 10.7.4.6). Many wall designs that are governed by slenderness effects also require compression forces in masonry coupling with reinforcement in tension and are therefore not affected. For example, single spans with only self-weight and some nominal loads transferred from roof level are often not loaded with enough axial load to move the location of the neutral axis beyond the depth of steel in the wall.

Additionally, in order for a design to be impacted by the presence of this issue, the load combinations where the circumstances above are present must also be critical to the design. Load combinations with the highest axial loads (for example: 1.4D) are most likely to also be governed by load combinations using high lateral loads combined AND lower axial loads (for example: 0.9D + 1.4W).

It might even be easier to rule out the wall designs that are not affected:

• all slender walls are unaffected and correctly handled within MASS
  
• all walls using vertical reinforcement in tension (bars are ignored in compression as they cannot be adequately tied) are unaffected
• all walls that do not have any slenderness effects are not affected
  

Most wall designs fall into one (or more) of these three categories and it takes a special combination of circumstances to trigger a design that results in this bug affecting software result

How to tell if a design is affected

Check to see if the steel is in tension

Look at the location of the neutral axis and compare it to where the vertical bars are placed. If the distance to the neutral axis, c, is less than the depth to the layer of reinforcement, d, then the steel is in tension and your design is not affected.

In this example, a wall constructed using 15cm units (thickness of 140mm) can be checked to see if the steel is in tension by comparing the Neutral axis value in the Simplified Moment Results window to the steel depth, d, which is placed in the middle of the wall (b_w/2).

Steel is in tension (c<d)

Check to see if that load case governs the design result

As mentioned earlier, the load cases resulting in the steel not being in tension tend to differ from the load cases that govern the final design result (1.4D compared to 1.25D or 0.9D + 1.4W). When this is the case, the design is unaffected by the result. The example file highlights this aspect where the load combination resulting in steel being ignored (L.C. #1) is not critical to the design. In this case, load combinations 2 and 3 have total factored moments that are closer to the envelope curve which are based on the correct critical compressive axial load.

Out-of-plane wall designs tend to be governed by load combinations that combine the largest lateral load with the lowest axial load which are more likely to include steel in tension.

Note: To recreate this example for yourself, design a 3m tall, simply supported wall using 15cm, 15MPa units. Apply an unfactored axial dead load of 60kN, turn off self-weight, and apply an unfactored uniformly distributed wind load of 1 kN/m (or 1 kPa when considering a 1m length of wall as is the case for MASS wall modules). When designing for moment and deflection, at the first thing that will happen is that the wall will correctly pass being designed as unreinforced. For the purposes of highlighting this bug, de-select the 0 bars/cell option, effectively forcing reinforcement to be placed in the wall and skipping any iterations that are reinforced.

If the wall design is affected, compare the M_r to the manually adjusted value of M_f,tot

If the bug is present in a MASS project, the total factored moment resistance can be adjusted using the following expression:

M_ftotadjusted represents the adjusted value for the total factored moment taking slenderness effects into account

P_cradjusted represents the adjusted critical axial compressive load

P_crMASS is the critical axial compressive load calculated by MASS for load combinations where the bug is present in the results.

All other terms are defined within the CSA S304-14.

Both a derivation of the P_cradjusted expression as well as an example demonstrating how to use it can be expanded in the two sections below by clicking on each heading:

It is expected that most designs will not meet all of the criteria for the bug to be present as the reinforcement is in tension for most reinforced wall designs. For those designs where the bug is present, it is possible for the the marginal difference in total factored moment to result in a failed design which would have been thought to have passed moment and deflection design. As mentioned earlier, the specific and unlikely combination of circumstances required to trigger this bug within MASS is the likely reason why it has remained undiscovered for so long.

Our Response

Bugs of this nature are taken very seriously. It was discovered in-house but not until very late in the Version 3.0 development process after it had been initially thought to be complete. As a result, the bug was investigated and a fix was added to Version 3.0 as well as to Version 2.2 in the form of MASS Version 2.2.1 (included in the installation directory for Version 3.0 – click here to read more). It has also been posted on our Known Bugs page where it links to this article.

If there is any question regarding the integrity of the results for a specific MASS project file, please feel free to contact CMDC directly. As the authorized MASS technical service provider, CMDC is available to help designers understand the specifics of identifying this issue, as well as any other masonry related technical questions. Click here for more information on technical assistance offered by CMDC.

This is one of the most common question we get having to do with masonry beam design so if this is something you are stuck on, know that you are not the first! We’ll first break down the reasons for this issue and then quickly outline what you can do about it.

First of all, this is not a bug in the code. This issue is a product of the way the 2004 S304 standard limits stirrup spacing. It has since been addressed in the 2014 edition and MASS Versions 3.0 and newer will allow two course designs with stirrups.

When does this happen?

This issue is specific to beams designs where:

1. The beam is 2 courses high or less, and
2. The beam is failing in shear design for having a shear resistance that is too low

You can see in the screenshot below that only the “none” option is selected for stirrup placement and if you try and click the single or double leg selections, nothing happens because the whole input area is greyed out or “disabled”.

The fix seems obvious, right? Just add some stirrups to boost my shear capacity! This is where we run into problems.

Why MASS won’t let you just add stirrups to your beam

The error message you are seeing reads: “Design fails: There is insufficient steel stirrup area and/or bar spacing according to CSA S304.1-04: 11.3.4.7.1,2 and/or 11.3.4.8”.

For reference, 11.3.4.7.1 and 11.3.4.7.2 each refer to stirrup placement and minimum area requirements. The issue we are running into comes from 11.3.4.8 which covers “Spacing limits for shear reinforcement”.

Click to see full reference

The issue here and the reason MASS will not let you place stirrups in your beam is that clause 11.3.4.8 is specifying a maximum spacing of d/2 which for 2 course beams is less than 200mm. MASS recognizes this and greys out the option to even place stirrups because the modular nature of masonry restricts you to multiples of 200mm (or the space between the centre of adjacent cells in concrete block construction).

Since you cannot physically place stirrups closer than 200mm within a masonry beam, MASS disable this option.

As you can see, there is no way to place stirrups any closer than one per cell. For those curious, these beams were built for university research supported by CMDC. You can read more about that here.

What you can do to get a successful design

You have a couple of options to get your design to pass. You can always send your MASS project file over to us in an email and we can walk through these options together over the phone as well. To do this, visit our contact page to get in touch.

1. Re-examine your loads

Oftentimes when going through a design, there are assumptions that are made along the way that can have a big impact on the final design. In the case of load distribution, are all of the loads being carried straight down to the beam or can arching be assumed as outlined on page 305 of our textbook? Did you manually add the self-weight of the beam in your dead load? If so, you might be double counting if you have the self-weight option selected in the loads window. Are there other loads applied that might not actually be resisted by the beam? All of these can mean the difference between a beam design passing and failing in shear design based on the loads applied.

2. Model an additional course of masonry within your beam

Many masonry beams are supporting more masonry above them so why not take advantage of this by considering a third course to be part of your beam? The real world difference could mean as little as simply grouting an additional course but the key difference is that the increase in height and in turn, d (depth of tension steel from compression face), moves the d/2 spacing restriction to being greater than 200mm. This might not be possible if the clear span is very small since the additional course might bump you into deep beam territory but for most designed masonry beams, adding the third course to your design does the trick!

3. Think about using a high lintel unit

If you can’t get your d/2 value below 200mm, you can approach the problem from the other direction and remove the 200mm restriction by changing to a high lintel beam design. High lintel beams have a continuous grouted area that can fit any reinforcement configuration. They can even be modeled in MASS for those of you who are feeling more ambitious! Give us a call if you are interested and we’ll be happy to walk you through it.

4. Consider using a stronger masonry unit

No, this does not mean you also need to increase your grout strength (See note 4 of Table 4 within the S304 masonry standard for reassurance). Using a unit with a higher f’m value increases the shear strength of the masonry itself, possibly giving you enough resistance as to no longer require stirrups in your beam. This is not always practical and is only effective if the failing masonry shear resistance is close to the required factored shear force.

One more thing…

Seeing as this whole issue is born not from a software bug but the way the CSA S304 standard was written, MASS Versions 3.0 and newer will not have this issue. CMDC currently maintains an active role in shaping and developing masonry standards and as a result, the 2014 edition of the S304 standard includes an addition to the “problem” clause, 11.3.4.8, which now includes a minimum value of 200mm for the maximum spacing of stirrups.

It is for this reason that the issue you are dealing with now will no longer be a problem when you start designing with the 2014 S304 standard.

Still have questions? Feel free to call or send us an email! (Including your MASS file is very helpful)