Perception and concentration: the story behind odour

Polyurethane foam is the major source of VOCs and odour in the automotive cockpit. Michael Holzwarth of design and development specialist imat-uve explains why it can be hard to change the smell of the foam

The global nature of the modern automotive industry makes cockpit odours a challenge. Those that are considered high status and pleasant smells in one part of the world can be perceived as unpleasant in another.

As car-makers try to produce one platform for each car worldwide, it is as important to deal with odours in the cab as it is to use uniform suspension or other physical components. This is why it is important to have a fundamental understanding of what smells are, and how humans perceive smell.

This has been an area of study for more than 100 years. But most of the knowledge generated by this branch of science is ignored by other sectors. We will use knowledge generated by that century of study to help explain the sense of smell and other sensory perceptions.

To evaluate the odour of a material, at least two questions must be answered. ‘What kind of odour is it?’ is the first, to establish odour quality. ‘What is the intensity of the odour?’ is the second, to determine odour quantity.

Strength and quantity

Before smell is tackled, it is perhaps simplest to approach the subject by building on more easily evaluated senses, starting with hearing. Sound waves are pressure fluctuations: volumes of air with higher and lower barometric pressure. Sound is mechanical energy that becomes a physical stimulus and, as is typical for physical phenomena, there is a mathematical description for sound. At its simplest, sound looks like a sine wave.

Waves have two key parameters: their frequency and their amplitude. The quality of sound is directly related to the pitch, and that relates to the frequency. The quantity of sound, or volume, is directly linked to the amplitude of the wave.

Physical measurements give us an idea of what the perception of the sound will be. This is quite simple, because there is a range of frequencies that humans can hear, and we know how sensitive human ears are in this range. It is possible to take a physical measurement, and this can be used to tell if a sound will be acceptable or not.

Next, taste, which is more similar to smell. Both rely on the chemical stimulation of sensory receptors. With taste, the receptors located on the tongue can distinguish between bitter, salty, sweet, umami, and sour. The intensity of a flavour depends on the number of receptors which it engages when it lands on the tongue.

A question of taste

It is comparatively easy to distinguish and describe tastes, because there are only five types of receptor. If a chemical does not stimulate a taste sensor, it will either stimulate nothing at all, or it will activate the sense of smell.

Tastes can only be measured qualitatively. For example, something is more or less salty than something else, but there is no established system to measure the intensity of taste.

The sense of smell also depends on chemicals stimulating sensors. These are located in our nasal mucosa. Smell is similar to taste, because quality is measured by the types of receptors, and the quantity or intensity depends on the number of receptors that we have. The main difference between smell and taste is complexity. Taste has five types of receptors; smell involves more than 300.

It is impossible to describe clearly in words what is perceived by 300 different receptors. We would have to find words for each of the receptors and all the possible mixtures. An odour is always a combination of all the stimulated receptors. This is why when six people smell an automotive interior, it is not unusual for them all to use different words to describe the smell.

The right language

Over the past 200 years, many people have tried to find a language and a systematic approach to enable a consistent description of an individual odour and its intensity. Many of these systems use completely different words to describe what the systems say are the same thing. Others use the same words, but the systems were developed independently in different locations at different times. It is not always clear that one use of a smell descriptor matches an earlier use of that same descriptor.

One approach takes chemical measurements to determine smell intensities. However, this tells us nothing about the odour as it does not link the mixtures to the smell people experience when they are exposed to that combination of chemicals.

Even if we know the chemical structure of an odour chemical, it gives no insight into which receptor is activated by the substance. Very similar substances can have completely different odours. Structural approaches do not work for either quality or quantity of smell.

Adding to the complexity, the sense of smell does not appear to be a continuous sense. Sound gives the impression of being continuous over different frequency ranges. But sound is a physical phenomenon detected in the ear. Smell is quite different – it is a chemical phenomenon that is recognised by cells that are sensitive to each smell at a different level.

Rating system

There are many established odour test methods, such as VDA 270, that allow participants to grade the intensity of odours in a subjective way. The VDA 270 test grades odours from 1 to 6, with half steps permitted. Grade 1 is not perceptible; Grade 2 is perceptible but not disturbing; and Grade 3 is clearly perceptible but not disturbing. Further up the scale, Grade 4 represents an odour that the participant deems disturbing; Grade 5 is strongly disturbing; and at the top of the scale, Grade 6 is an odour the participant considers not acceptable. This scale is used in figures 1 and 2. In figure 2: 1-3 are green, 3.5 is yellow and 4-6 are red.

However, this system is not without its problems. An odour one person may find clearly perceptible but not disturbing might be considered disturbing by someone else. The results rely on intensity, and whether people like the smell or not.

When trying to look in more detail into the perception of odour intensities, a good starting point is the work of the 19th century German physician Ernst Heinrich Weber, who was one of the developers of psychophysics. This discipline studies the relationships between an objective stimulus and the subject’s perception of that stimulus.

The differential perception threshold is a very important concept in psychophysics, and comes directly from Weber’s pioneering work. He discovered that the difference between two stimuli has to exceed a certain value if human senses are to be able discern the difference.

For example, a human would not be able to tell the difference between two weights of 400g and 401g. But, if they were given two weights of 100g and 102g, and the conditions are right, many people can tell the difference between the two. The number will be less than 100% of trialists, but it will be much higher than 50%.

Weber and his student Gustav Fechner found that, typically, the minimum perceptible difference for weight is 2%. For taste, it is between 10-30%, the same as smell. Down to a quarter

Because the range of difference needed to guarantee the perception of a different smell intensity is between 10 and 30%, we work on the basis that the smell level must change by at least 25%. This simple experiment demonstrates the consequence of Weber and Fechner’s work.

A test subject is exposed to two identical stimuli of a perceivable intensity. One of the stimuli is steadily increased, until the subject notices the difference.

This is step 1 in figure 1, where the x-axis shows the intensity of the stimulus and the y-axis how strongly that stimulus is perceived.

After adjusting the both of the stimuli to the higher level at the end of step 1, one stimulus is increased until that increase is perceptible.

Weber’s work indicates that the increase of stimulation in the second step, which starts at a higher level, has to be higher than in the first step. But the difference perception of the test subject in both cases has the same extent. The difference in both cases was just noticeable.

The stimulus is increased further, and responses noted until the subject cannot detect any more changes. If all of these results are plotted on a graph, the logarithmic relationship between the perception of an odour’s intensity and the actual concentration becomes clear. This is one the reasons why there are so many misunderstandings and so much frustration in the world of odour.

Frustration

Moving to the world of automotive moulding, take an automotive seat foam cushion that has received an odour evaluation grade 5.5 on our virtual scale of perceived odour intensity. The cushion maker is in trouble, because the car-maker will not accept the smell of the cushion.

In this situation, the cushion maker will typically do what they can to remove the smell. They will improve the recipe, and they may change the production process. If they manage to reduce VOCs to 50% of the original level, this may lead them to believe that the odour will also have improved. However, they will often be frustrated to find that in odour tests, participants are barely able to discern any significant difference between the odour of the old cushions and the new ones.

This is because a 50% reduction in stimulus produces a far lower reduction in perception at the on a logarithmic curve. The difference of perception here is not the difference in stimulus.

To improve the odour perception to maybe 50% of the original level, the quantity of odorous chemical may have to fall. This could be from 100% of its original value to 10% of the original value before people are able perceive a significant difference.

Therefore, as a rule of thumb, for a significant reduction in the perceived odour intensity, the stimulus should be reduced to 10% of the original amount. Otherwise, there is no guarantee that people will notice the improvement.

Rule of thumb

Figure 2 shows how the VDA 270 scale could look in model cases. The graph shows a range of concentrations for four specific chemicals. Two of them, A and C, are strong smells and their concentration/perception curves are almost vertical. The smells of B and D are much weaker, and their response curves are more horizontal in shape.

This shows that B and D can be present at relatively high concentrations with few problems.

But each of these curves is for an individual chemical, and the smell of the object is made up not only of each of these odours, but it is influenced by the way that the different odours interact.

That said, in this graph, samples C and D are not causing any problems as they are at, or below, the acceptable VDA 720 Grade 3 level. To reduce the odour, work needs to be done to reduce the perceptions of both B and A.

The goal is for sample B is to come down from Grade 5 to Grade 3, and for sample A move from Grade 5.5 down to Grade 3.

For sample A, this can be done by reducing its concentration from 9 to 3, a reduction of 66%. For sample B, the concentration must fall from 96 to 16, a decrease of more than 80%.

Of course, it may not be possible to reduce the concentration of B that much if it is only present at a low level, because that could lead to an unsuccessful formulation. That’s where the skill of the polyurethane cushion maker is called for..

* Michael Holzwarth is senior manager for VIAQ, VOC and odour at imat-uve. This is an edited version of a paper presented at the CPI Meeting held in Atlanta, Georgia in late September 2018.