Material Blending

Blending material gradations to obtain consistent characteristics and performance can be a tremendous challenge, especially when faced with sources that can change over time.  Without tools to show you both mathematically and graphically what is happening, means that any adjustments to the blend are like a shot in the dark.  Not only can this become an endless cycle of trial and error, it can be costly.
But blending materials is not just about maintaining consistency, it’s also about optimizing blends to meet specific requirements.  That requirement may be to produce a concrete sand blend from different sand classifier stations; it may be the optimum combination of materials needed to produce the lowest cost blend that still meets the specification; a gradation that meets a specific job-mix formula (JMF) and uses a fixed amount of RAP; it may be the densest blend of materials needed for an asphalt mix; or it could be the closest blend of materials needed to make a good pumpable concrete mix.  A blending tool needs to be robust enough to allow for multiple methods to adjust, evaluate, and optimize material blends.  Stonemont Software blending tools not only meet those requirements, our automated features give the user time to really focus on the results rather than the process.
Stonemont Software includes blending tools for aggregate, asphalt, and concrete blending.  Many of the tools are the same but there are a few specific blending tools specific to concrete mixes.  All the blending tools within Stonemont Software have a similar and familiar layout.  They allow the user to hand-enter or query in component materials, and enter or load specifications and targets.
Historically, blending was performed by manually changing the percent contribution of each component material until a satisfactory blend was achieved.  This manual blending technique is supported in Stonemont Software and is still useful if there are fixed percentages that need to be met or specific ratios of materials that are required.  However, manual blending can be very frustrating and time-consuming and typically those solutions would not be considered optimal.  Therefore, Stonemont Software offers the unsurpassed benefits of manual, mathematical, and visual blending.
Mathematical Blending
With the wide use of spreadsheets came mathematical blending, which allowed users to save significant time by reducing the iterations necessary by manual blending.  Mathematical blending allows more complexity to be introduced to the blending process by adding an objective function and constraints to the problem.  For example, if you are trying to produce the lowest cost blend that meets a certain specification or target, then lowest cost is the objective function and the specification or target values are mathematical constraints on the blending process.  The objective function doesn’t have to be in terms of price, it can be thought of as percentage of material.  For example, if you have low inventory on a particular material that is used in an existing blend, you can set the objective function value for that material to be lower than all other materials in the blend and maximize the blend objective rather than minimize the objective.
Figure 1.

The most common constraint is to include specifications or target values (Figure 1).  Specifications are typically an external constraint prescribed by an agency or customer.  Targets are typically an in internal constraint used to improve process control and provide a warning band prior to exceeding specifications.   Specifications and targets are usually a band including an upper and lower value.  A target or JMF is typically a single constraint from which a specification or target band can be developed.  Blending to a target or JMF is useful when attempting to meet a prescribed blend or recreate a blend, possibly using different materials, that was particularly suitable for a given purpose.  As previously mentioned, blending to specifications is useful when trying to minimize or maximize the objective function of the blend.  However, you must be careful to not use actual prescribed specifications since mathematical blending may cause the blended result of at least one sieve or other quality parameter to be on or near a specification limit.  Therefore, it is recommended to adjust your specifications to provide a margin of safety.  The MixRisk tool that was discussed previously can help ensure that your blend specifications are appropriate and help you better understand the potential future performance of the blend. 

Sometimes, cost is not the overriding factor but rather to produce a blend that will reduce the risk of future performance issues.  That ideal or least risk blend would be one that is centered within the specifications.  Stonemont Software provides a tool to quickly blend to the middle of the specifications, which reduces the guess work and manual iterations necessary to achieve such a blend.  This tool is often used to get the blend as close to ideal as possible before further optimizing the blend for a particular need.  It also is a quick way to see how different blends of materials can be modified to meet different specifications.  Sometimes the priority is to create a densely graded asphalt or concrete mix.  Stonemont Software provides the ability to quickly blend to the maximum density line on a .45 power chart.  It is used often as a toggle to see the difference between blending to meet a specification and blending to max-density.
A common constraint includes specifying a fixed percentage of component material to use in the blend.  In an aggregate blend you may need to maintain a specific percentage of a material to balance the plant, in an asphalt mix you may want to fix the percentage of RAP in the blend, for a concrete mix you will likely choose to lock the cement in the mix while optimizing the rest of the blend.  Another constraint is to specify minimum and/or maximum percent contributions for each component material in the blend.   This constraint can help prevent one of the unwelcome artifacts of mathematical blending, which is setting material contributions to unpractical values or not setting them at all since the blend constraints can be met without contribution from some materials.  For example, a fine sand bin may plug up if the percentage of material is too small to allow the gates to open up wide enough for the material to flow.  When batching materials, too much of one material may limit how much can be batched at one time.  Too little material can create excessively long batches while the plant attempts to jog the small amount or worse, drops too much because it cannot batch such a small amount.  These are but a few of the reasons why it may be practical to put constraints on the minimum and maximum component percentages in a blend. 
As more constraints are added or the tighter the constraints are on the blending process the more control they exert on the solution.   For example, if specifications are entered for all sieves to be 0 to 100 then the objective function to minimize cost will control the solution.  As we tighten the specifications they will limit the solution to an acceptable range.  An important advantage of mathematical blending over manual blending is that you will be notified immediately if no solution is possible given the current constraints, which save a lot of time manually iterating material percentages before realizing no solution is possible.
Visual Bending
There are times when you may want to modify a manual or mathematically produced blend.  One example may be to move the resulting gradation further from a specification boundary.  Sometimes it is easier to do this visually rather than in tabular form, so Stonemont Software offers a very powerful suite of visual blending tools.  These tools allow users to point-and-click or drag-and-drop points on charts to specify new target blending values.
Sieve Charts
With the sieve chart visual blending tool, it is simply a matter of dragging-and-dropping target gradation values and then recalculating the blend to see how close the materials can actually get to that blend.  This is a great tool for quickly achieving a reasonable compromise with less than optimum materials or for fine tuning a blend that was started using some of the other tools.  In the example shown in Figure 2, targets (green triangles) were dragged-and-dropped from the original locations (open circles) with the intention to fine up the blend on the coarser sieves and coarsen up the blend on the finer sieves.   The results shown in Figure 3 are a close match to the target gradation.  If satisfied with the results, they can be posted to the material blend or targets can be re dragged and another attempt can be made to further modify the blend.  


Figure 2.
Figure 3.


Power Charts
The power chart visual blending tool functions similarly to the sieve chart blending tool but provides the .45 power curve as the canvas for drag-and-drop functionality.  This tool will calculate the max density line based on the current gradation.  If changes to the gradation are significant enough to affect the max density line then it will automatically shift according to the updated gradation.  This tool is very useful for fine tuning a blend of materials.  For this example, the goal was to move the blend closer to the maximum density line (but not exactly to the maximum density line or we would use the blend to maximum density line functionality!).  Figure 4 shows how the targets (green triangles) were moved closer to the maximum density line from their original locations (black triangles).  Figure 5 shows the results of calculating the blend using the new visually created targets.  As you can see the fit is not perfect to the target locations.  When working with material blends, including several component materials can really help with fine tuning and improves the chances that a given blend can be met.  What works best may surprise you.
Figure 4.
Figure 5.
Individual Retained Charts
What works best for percent passing specifications does not always work out best when considering individual retained specifications.  The individual retained chart blending tool is available for those times when meeting an individual retained specification is necessary.  Primarily found in concrete specifications, the individual retained values for a material blend can give a better understanding of how the material is distributed throughout the gradation.  Spikes in material retained on any one sieve are quite apparent on an individual retained chart which can often go unnoticed on a percent passing chart.
In this example, the original material blend (blue diamonds) has a significant portion retained on the #100 sieve (Figure 6).  Since this is a fine aggregate material blend from the combination of classifier stations, the target gradation (green triangles) has been dragged out to fit a haystack configuration.  Figure 7 demonstrates how well blending can be fine-tuned when several materials are considered as part of the blend.
Figure 6.
Figure 7.
Coarseness Workability Charts
Unique to concrete mix designs, the coarseness workability visual blending tool is a good starting point when adjusting a concrete mix.  When combined with the other visual blending and evaluation tools, it can give you a well-rounded view of how a concrete mix may potentially perform.  It is important to note that this tool is only used to modify the aggregate proportion of the concrete blend so the adjusted workability value is not shown on this chart.  The adjusted workability value can be viewed on our standard coarseness workability charts for concrete mix design development that are not used for visual blending.
In this example, the goal was to slightly improve the workability of the mix.  The blue dot in Figure 8 represents the blended material in the original concrete mix.  To adjust the mix and improve the workability according to the coarseness workability chart, simply click an alternative target location as shown on Figure 9.  If the new aggregate blend is possible, the adjustment can be posted back to the mix and all material amounts and percentages will be updated accordingly.  The program will provide immediate feedback if it is not possible to achieve the attempted adjustment.  See a short video of visual blending using the coarseness workability chart in Stonemont Software.  Other concrete specific blending tools include the ability to blend to a specific unit weight or density.
Figure 8.
Figure 9.
Although spreadsheets allow for mathematical blending, their usefulness is limited because they are typically not an integrated part of a quality control and mix design software package.  Stonemont Software offers the unsurpassed benefits of manual, mathematical, and visual blending.  Stonemont Software also offers the ability to manage all blends company-wide and the ability to easily query for changes in component materials that makes for a very robust set of blending tools. 

Material blending is a critical need for the aggregate, asphalt, and concrete material industries.  If you are wasting time exporting material gradations to a spreadsheet to perform material blending or your mix design software does not provide adequate blending tools then it is time to step up to the powerful blending tools that are an integral part of the Stonemont quality control and mix design software. 

For more information please contact Stonemont Solutions, Inc. 

Michael Rodriguez
Adrian Field
Stonemont Solutions, Inc.